Sunday


Wilhelm Dilthey (19 November 1833 to 01 October 1911)

Every so often a term from philosophy — and by “philosophy” in this context I mean the kind of philosophy that is generally not read by the wider public, and which is therefore sometimes called “technical” or “professional” — finds its way into the wild, as it were, and begins to appear in non-philosophical contexts. This happened with Thomas Kuhn’s use of “paradigm shift” and with Derrida’s use of “deconstruction.” To a lesser extent, it is also true of “phenomenology” since Husserl’s use of the term. Another philosophical term that has come into general currency is “lived experience.” (There are also variations on the theme of “lived experience,” such as “felt experience,” which I found in Barry Mazur’s 2008 paper “Mathematical Platonism and its Opposites,” in which the author refers to, “…the passionate felt experience that makes it so wonderful to think mathematics.”) Recently I saw “lived experience” used in the title of a non-philosophical book, Nubia in the New Kingdom: Lived Experience, Pharaonic Control and Indigenous Traditions, edited by N. Spencer, A. Stevens, and M. Binder. A description of the book on the publishers website says that the approach of the volume provides, “…a more nuanced understanding of what it was like to live in colonial Kush during the later second millennium BC.”

This, I think, is the takeaway of “lived experience” for non-philosophers — that of “what it was like to live” in some particular social or historical context. One could easily imagine, “what it was really like to live” becoming a slogan on a par with Leopold von Ranke’s, “to show what actually happened” (“wie es eigentlich gewesen”). Both could be taken as historiographical principles, and indeed the two might be taken to imply each other: arguably, one can’t know what it was like to live without knowing what actually happened, and, again arguably, one can’t show what actually happened without knowing what it was like to live. Actually, I think that the two are distinguishable, but I only wanted to make the point of how closely related these ideas are.

I believe, though I cannot say for sure, that the philosophical use of “lived experience” originates in the work of Wilhelm Dilthey. If Dilthey did not originate the philosophical use of “lived experience,” he did write extensively about it earlier than most other philosophers who took up the term. (If anyone knows otherwise, please set me straight.) Since I am planning on making use of the idea of lived experience, I have been reading Dilthey recently, especially his Selected Works, Volume III: The Formation of the Historical World in the Human Sciences (which corresponds to the German language Gesammelte Schriften, Volume 7: Der Aufbau der geschichtlichen Welt in den Geisteswissenschaften), which has a lot of material on lived experience.

Dilthey is not an easy author to read. I have heard it said many times that Husserl is a difficult author, but I find translations of Husserl to be much easier going than translations of Dilthey. Dilthey and Husserl knew each other, read each others’ works, and they corresponded. Dilthey’s exposition of lived experience contains numerous references to Husserl’s Logical Investigations (Husserl’s systematic works on phenomenology mostly appeared after Dilthey passed away, so it was only the Logical Investigations to which Dilthey had access). Most interestingly to me, Husserl wrote a semi-polemical article, “Philosophy as Rigorous Science,” in which Husserl discussed Dilthey in the section “Historicism and Weltanschauung Philosophy.” Dilthey did not agree with the characterization of his work by Husserl. It was Husserl’s article that was the occasion of their correspondence (translated in Husserl: Shorter Works), and it is a lesson in the unity German philosophy to read this exchange of letters. In their correspondence, Dilthey and Husserl were easily able to find common ground in a language rooted in 19th century German idealist philosophy.

While the apparent ground of their common outlook was expressed in the peculiar idiom of German philosophy, both were also reacting against that tradition. Both Dilthey and Husserl were centrally concerned with the experience of time. Husserl’s manuscripts on time consciousness run to hundreds of pages (cf. On the Phenomenology of the Consciousness of Internal Time (1893–1917)). Of Husserl’s efforts Dilthey wrote, “A true Plato, who first of all fixes in concept the things that become and flow, then puts beside the concept of the fixed a concept of flowing.” (cited by Quentin Lauer in The Triumph of Subjectivity from Dilthey, Gesammelte Schriften, Vol. V, p. cxii) Dilthey’s own exposition of time consciousness can be found in Vol. III of the selected works in English, Drafts for a Critique of Historical Reason, section 2, “Reflexive Awareness, Reality: Time” (pp. 214-218), where it is integral with his exposition of lived experience.

Of time and lived experience Dilthey wrote:

“Temporality is contained in life as its first categorical determination and the one that is fundamental for all others… Thus the lived experience of time determines the content of our lives in all directions.”

Wilhem Dilthey, Selected Works, Volume III: The Formation of the Historical World in the Human Sciences, Princeton and Oxford: Princeton University Press, 2002, pp. 214-215.

I suspect that Husserl would have agreed with this, as for Husserl time consciousness was the foundation of the constituting consciousness. Dilthey also writes:

“That which forms a unity of presence in the flow of time because it has a unitary meaning is the smallest unit definable as a lived experience.” And, “A lived experience is a temporal sequence in which every state is a flux before it can become a distinct object.” And, “The course of life consists of parts, of lived experiences that are inwardly connected with each other. Each lived experience relates to a self of which it is a part.”

Op. cit., pp. 216-217

Here I have plucked out a few representative quotes by Dilthey on lived experience; this may give a flavor of his exposition, but I certainly don’t maintain that this is a fair way of coming to grips with Dilthey’s conception of lived experience. The only way to do that is by the lived experience of reading the text through and deriving from it a unitary meaning. I will not attempt to do that in the present context, as I only wanted here to give the reader an impression of Dilthey’s writing on lived experience.

Dilthey, as I noted, is not an easy author. Both Dilthey’s and Husserl’s discussions of time consciousness and lived experience are opaque at best. I keep at Dilthey despite the difficulty because I want to understand his exposition of lived experience. However, as I keep at it I cannot help but think that part of the difficulty of the discussion is the absence of a scientific understanding of consciousness. As I have mentioned many times, we simply have no idea, at the present stage of the development of our scientific knowledge, what consciousness is. Trying to give a detailed description of time consciousness and lived experience without any scientific foundation is almost crippling. I believe that the effort is worthwhile, but it is as instructive in how it fails as it is instructive in how it less often succeeds.

In this frame of mind I recalled a passage from Foucault’s The Birth of the Clinic:

“Towards the middle of the eighteenth century, Pomme treated and cured a hysteric by making her take ‘baths, ten or twelve hours a day, for ten whole months.’ At the end of this treatment for the desiccation of the nervous system and the heat that sustained it, Pomme saw ‘membranous tissues like pieces of damp parchment … peel away with some slight discomfort, and these were passed daily with the urine; the right ureter also peeled away and came out whole in the same way.’ The same thing occurred with the intestines, which at another stage, ‘peeled off their internal tunics, which we saw emerge from the rectum. The oesophagus, the arterial trachea, and the tongue also peeled in due course; and the patient had rejected different pieces either by vomiting or by expectoration’.”

“…Pomme, lacking any perceptual base, speaks to us in the language of fantasy. But by what fundamental experience can we establish such an obvious difference below the level of our certainties, in that region from which they emerge? How can we be sure that an eighteenth-century doctor did not see what he saw, but that it needed several decades before the fantastic figures were dissipated to reveal, in the space they vacated, the shapes of things as they really are?”

Michel Foucault, The Birth of the Clinic, New York: Vintage, 1975, pp. ix-x; Foucault cites Pomme, Traite des affections vaporeuses des deux sexes (4th edn., Lyons, 1769, vol. I, pp. 60-5)

Because of the theory-ladenness of perception, when the theory is absent or unclear, perception has little to go on and it is confused and unclear. We cannot describe with precision unless we can conceptualize with precision. The eventual development of an adequate science of consciousness — which may ultimately involve a revision to the nature of science itself — will issue in concepts of sufficient precision that they can be the basis of precise observations, and precise observations can further contribute to the precisification of the concepts — a virtuous circle of expanding knowledge.

I would not insist upon the theory-ladenness of perception to the point of excluding the possibility of any knowledge without an adequate theory to guide perception. In this spirit I have already acknowledged that there is some value in Dilthey’s attempt to clarify the idea of lived experience. If theory and observation are mutually implicated, and eventually can accelerate in a virtuous circle of mutual clarification, then the first, tentative ideas and observations on lived experience can be understood analogously to the stone tools used by our earliest ancestors. These stone tools are rough and rudimentary by present standards of precision machine tools, but we had to start somewhere. So too with our conceptual tools: we have to start somewhere.

Dilthey’s approach to lived experience is one such starting point, and from this point of departure we can revise, amend, and extend Dilthey’s conception until it becomes a more useful tool for us. One way to do this is by way of what has been called the knowledge argument, also known as the Mary’s room thought experiment. I have earlier discussed the knowledge argument in Colonia del Sacramento and the Knowledge Argument and Computational Omniscience.

Here is the locus classicus of the thought experiment:

“Mary is a brilliant scientist who is, for whatever reason, forced to investigate the world from a black and white room via a black and white television monitor. She specializes in the neurophysiology of vision and acquires, let us suppose, all the physical information there is to obtain about what goes on when we see ripe tomatoes, or the sky, and use terms like ‘red,’ ‘blue,’ and so on. She discovers, for example, just which wavelength combinations from the sky stimulate the retina, and exactly how this produces via the central nervous system the contraction of the vocal cords and expulsion of air from the lungs that results in the uttering of the sentence ‘The sky is blue.’ […] What will happen when Mary is released from her black and white room or is given a color television monitor? Will she learn anything or not?”

Frank Jackson, “Epiphenomenal Qualia” (1982)

The historical parallel of the Mary’s room argument would be to ask, if Mary had exhaustively studied life in colonial Kush during the later second millennium BC, and then Mary was enabled to actually go back and live in colonial Kush during the later second millennium BC, would Mary learn anything by the latter method that she did not already know from the first method? If we answer that Mary learns nothing from living in Kush that she did not already know by exhaustively studying Kush, then we can assert the equivalence of what it was like to live and what actually happened. If, on the other hand, we answer that Mary does indeed learn something from living in Kush that she did not learn by exhaustively studying Kush, then we ought to deny the equivalence of what it was like to live and what actually happened.

While this exact thought experiment cannot be performed, there is a more mundane parallel that anyone can test: exhaustively educate yourself about somewhere you have never visited, and then go to see the place for yourself. Do you learn anything when you visit that you did not know from your prior exhaustive study? In other words, does the lived experience of the place add to the knowledge you had gained without lived experience?

While Dilthey does not use the term “ineffable,” many of his formulations of lived experience point to its ineffability and our inability to capture lived experience in any conceptual framework (as is implied by his criticism of Husserl, quoted above). If what one learns from what it was like to live is ineffable, then we could assert that, even when our conceptual framework was as adequate as we can make it, it is still inadequate and leaves out something of what what it was like to live, i.e., it leaves out the component of lived experience.

But, as I said, Dilthey himself does not use the term “ineffable” in this context, and he may have avoided it for the best scientific reasons. Our inability to formulate the distinctiveness of lived experience in contradistinction to that which can be learned apart from lived experience may be simply due to the inadequacy of our conceptual framework. When we have improved our conceptual framework, we may possess the concepts necessary to render that which now appears ineffable as something that can be accounted for in our conceptual framework. We must admit in all honesty, however, that we aren’t there yet in relation to lived experience. This is not a reason to avoid the concept of lived experience, but, on the contrary, it is a reason to work all the more diligently at clarifying the concept of lived experience. Employing simple distinctions like that between what it was like to live and what actually happened is one way to test the boundaries of the concept and so to better understand its relationships to other related concepts.

. . . . .

Pierre Pomme (1735 to 1812)

. . . . .

signature

. . . . .

Grand Strategy Annex

. . . . .

project astrolabe logo smaller

. . . . .

Advertisements

Computational Omniscience

18 December 2013

Wednesday


hal9000

What does it mean for a body of knowledge to be founded in fact? This is a central question in the philosophy of science: do the facts suggest hypotheses, or are hypotheses used to give meanings to facts? These questions are also posed in history and the philosophy of history. Is history a body of knowledge founded on facts? What else could it be? But do the facts of history suggest historical hypotheses, or do our historical hypotheses give meaning and value to historical facts, without which the bare facts would add up to nothing?

Is history a science? Can we analyze the body of historical knowledge in terms of facts and hypotheses? Is history subject to the same constraints and possibilities as science? An hypothesis is an opportunity — an opportunity to transform facts in the image of meaning; facts are limitations that constrain hypotheses. An hypothesis is an epistemic opportunity — an opportunity to make sense of the world — and therefore an hypothesis is also at the same time an epistemic risk — a risk of getting interpreting the world incorrectly and misunderstanding events.

The ancient question of whether history is an art or a science would seem to have been settled by the emergence of scientific historiography, which clearly is a science, but this does not answer the question of what history was before scientific historiography. One might reasonably maintain that scientific historiography was the implicit telos of all previous historiographical study, but this fails to acknowledge the role of historical narratives in shaping our multiple human identities — personal, cultural, ethnic, political, mythological.

If Big History should become the basis of some future axialization of industrial-technological civilization, then scientific historiography too will play a constitutive role in human identity, and while other and older identity narratives presently coexist and furnish different individuals with a different sense of their place in the world, we have already seen the beginnings of an identity shaped by science.

There is a sense in which the scientific historian today knows much more about the past than those who lived in the past and experienced that past as an immediate yet fragmentary present. One might infer the possibility of a total knowledge of the past through the cumulative knowledge scientific historiography — a condition denied to those who actually lived in the past — although this “total” knowledge must fall short of the peculiar kind of knowledge derived from immediate personal experience, as contemplated in the thought experiment known as “Mary’s room.”

In the thought experiment known as Mary’s room, also called the knowledge argument, we imagine a condition of total knowledge and compare this with the peculiar kind of knowledge that is derived from experience, in contradistinction to the knowledge of knowledge we come to through science. Here is the source of the Mary’s room thought experiment:

“Mary is a brilliant scientist who is, for whatever reason, forced to investigate the world from a black and white room via a black and white television monitor. She specializes in the neurophysiology of vision and acquires, let us suppose, all the physical information there is to obtain about what goes on when we see ripe tomatoes, or the sky, and use terms like ‘red’, ‘blue’, and so on. She discovers, for example, just which wavelength combinations from the sky stimulate the retina, and exactly how this produces via the central nervous system the contraction of the vocal cords and expulsion of air from the lungs that results in the uttering of the sentence ‘The sky is blue’. […] What will happen when Mary is released from her black and white room or is given a color television monitor? Will she learn anything or not?”

Frank Jackson, “Epiphenomenal Qualia” (1982)

Philosophers disagree on whether Mary learns anything upon leaving Mary’s room. As a thought experiment, it is intended not to give as a definitive answer to a circumstance that is never likely to occur in fact, but to sharpen our intuitions and refine our formulations. We can try to do the same with formulations of an ideal totality of knowledge derived from scientific historiography. There is a sense in which scientific historiography allows us to know much more about the past than those who lived in the past. To echo a question of Thomas Nagel, was there something that it was like to be in the past? Are there, or were there, historical qualia? Is the total knowledge of history afforded by scientific historiography short of capturing historical qualia?

In the Mary’s room thought experiment the agent in question is human and the experience is imposed colorblindness. Many people live with colorblindness within the condition greatly impacting their lives, so in this context it is plausible that Mary learns nothing upon the lifting of her imposed colorblindness, since the gap between these conditions is not as intuitively obvious as the gap between agents of a fundamentally different kind (as, e.g., distinct species) or between experiences of a fundamental different kind in which it is not plausible that the the lifting of an imposed limitation on experience results in no significant impact on one’s life.

We can sharpen the formulation of Mary’s room, and thus potentially sharpen our own intuitions, by taking a more intense experience than that of color vision. We can also alter the sense of this thought experiment by considering the question across distinct species or across the division between minds and machines. For example, if a machine learned everything that there is to know about eating would that machine know what it was like to eat? Would total knowledge after the manner of Mary’s knowledge of color suffice to exhaust knowledge of eating, even in the absence of an actual experience of eating? I doubt that many would be convinced that learning about eating without the experience of eating would be sufficient to exhaust what there is to know about eating. Thomas Nagel’s thought experiment in “What is it like to be a bat?” alluded to above poses the knowledge argument across species.

We can give this same thought experiment yet another twist if we reverse the roles of minds and machines, and asking of machine experience, should machine consciousness emerge, the questions we have asked of human experience (or bat experience). If a human being learned everything there is to know about AI and machine consciousness, would such a human being know what it is like to be a machine? Could knowledge of machines exhaust uniquely machine experience?

The kind of total scientific knowledge of the world implicit in scientific historiography is not unlike what Pierre Simon LaPlace had in mind when he posited the possibility of predicting the entire state of the universe, past or future, on the basis of an exhaustive knowledge of the present. LaPlace’s argument is also a classic determinist position:

“We ought then to regard the present state of the universe as the effect of its anterior state and as the cause of the one which is to follow. Given for one instant an intelligence which could comprehend all the forces by which nature is animated and the respective situation of the beings who compose it — an intelligence sufficiently vast to submit these data to analysis — it would embrace in the same formula the movements of the greatest bodies of the universe and those of the lightest atom; for it, nothing would be uncertain and the future, as the past, would be present to its eyes. The human mind offers, in the perfection which it has been able to give to astronomy, a feeble idea of this intelligence. Its discoveries in mechanics and geometry, added to that of universal gravity, have enabled it to comprehend in the same analytical expressions the past and future states of the system of the world. Applying the same method to some other objects of its knowledge, it has succeeded in referring to general laws observed phenomena and in foreseeing those which given circumstances ought to produce. All these efforts in the search for truth tend to lead it back continually to the vast intelligence which we have just mentioned, but from which it will always remain infinitely removed. This tendency, peculiar to the human race, is that which renders it superior to animals; and their progress in this respect distinguishes nations and ages and constitutes their true glory.”

A Philosophical Essay on Probabilities, PIERRE SIMON, MARQUIS DE LAPLACE, WITH AN INTRODUCTORY NOTE BY E. T. BELL, DOVER PUBLICATIONS, INC., New York, CHAPTER II.

While such a LaPlacean calculation of the universe would lie beyond the capability of any human being, it might someday lie within the capacity of another kind of intelligence. What LaPlace here calls, “an intelligence sufficiently vast to submit these data to analysis,” suggests the possibility of a sufficiently advanced (i.e., sufficiently large and fast) computer that could make this calculation, thereby achieving a kind of computational omniscience.

Long before we have reached the point of an “intelligence explosion” (the storied “technological singularity”) and machines surpass the intelligence of human beings, and each generation of machine is able to build a yet more intelligent successor (i.e., an “intelligence explosion”), the computational power at our disposal will for all practical purposes exhaust the world and will thus have obtained computational omniscience. We have already begun to converge upon this kind of total knowledge of the cosmos with the Bolshoi Cosmological Simulations and similar efforts with other supercomputers.

It is this kind of reasoning in regard to the future of cosmological simulations that has led to contemporary formulations of the “Simulation Hypothesis” — the hypothesis that we are, ourselves, at this moment, living in a computer simulation. According to the simulation argument, cosmological simulations become so elaborate and are refined to such a fine-grained level of detail that the simulation eventually populates itself with conscious agents, i.e., ourselves. Here, the map really does coincide with the territory, at least for us. The entity or entities conducting such a grand simulation, and presumably standing outside the whole simulation observing, can see the simulation for the simulation that it is. (The connection between cosmology and the simulation argument is nicely explained in the episode “Are We Real?” of the television series “What We Still Don’t Know” hosted by noted cosmologist Martin Rees.)

One way to formulate the idea of omniscience is to define omniscience as knowledge of the absolute infinite. The absolute infinite is an inconsistent multiplicity (in Cantorian terms). There is a certain reasonableness in this, as the logical principle of explosion, also known as ex falso quodlibet (namely, the principle that anything follows from a contradiction), means that an inconsistent multiplicity that incorporates contradictions is far richer than any consistent multiplicity. In so far as omniscience could be defined as knowledge of the absolute infinite, few would, I think, be willing to argue for the possibility of computational omniscience, so we will below pursue this from another angle, but I wanted to mention this idea of defining omniscience as knowledge of the absolute infinite because it strikes me as interesting. But no more of this for now.

The claim of computational omniscience must be qualified, since computational omniscience can exhaust only that portion of the world exhaustible by computational means; computational omniscience is the kind of omniscience that we encountered in the “Mary’s room” thought experiment, which might plausibly be thought to exhaust the world, or which might with equal plausibility be seen as falling far short of all that might be known of some body of knowledge.

Computational omniscience is distinct from omniscience simpliciter; while exhaustive in one respect, it fails to capture certain aspects of the world. Computational omniscience may be defined as the computation of all that is potentially computable, which leaves aside that which is not computable. The non-computable aspects of the world include, but are not limited to, non-computable functions, quantum indeterminacy, that which is non-quantifiable (for whatever reason), the qualitative dimension of conscious experience (i.e., qualia), and that which is inferred but not observable. These are pretty significant exceptions. What is left over? What part of the world is computable? This is a philosophical question that we must ask once we understand that computability has limits and that these limits may be distinct from the limits of human intelligence. Just as conscious biological agents face intrinsic epistemic limits, so also non-biological agents would also face intrinsic epistemic limits — in so far as a non-biological agent can be considered an epistemic agent — but these limitations on biological and non-biological agents are not necessarily the same.

The ultimate inadequacy of computational omniscience points to the possibility of limited omniscience — though one might well assert that omniscience that is limited is not really omniscience at all. The limited omniscience of a computer capable of computing the fate of the known universe may be compared to recent research on what Daniel Kahneman calls the bounded rationality of human minds. Artificial intelligence is likely to be a bounded intelligence that exemplifies bounded rationality, although its boundaries will not necessarily coincide precisely with the boundaries that defined human bounded rationality.

The idea of limited omniscience has been explored in mathematics, particular in regard to constructivism. Constructivist mathematicians have formulated principles of omniscience, and, wary of both unrestricted use of tertium non datur and of its complete interdiction in the manner of intuitionism, the limited principle of omniscience has been proposed as a specific way to skirt around some of the problems implicit in the realism of unrestricted tertium non datur.

When we allow our mathematical thought to coincide with realities and infinities — an approach that we are assured is practical and empirical, and bound to only yield benefits — we find ourselves mired in paradoxes, and in the interest of freeing ourselves from this conceptual mire we are driven to a position like Einstein’s famous aphorism that, “As far as the laws of mathematics refer to reality, they are not certain; and as far as they are certain, they do not refer to reality.” We separate and compartmentalize factual realities and mathematical infinities because we have difficulty, “to hold two opposing ideas in mind at the same time and still retain the ability to function.”

Indeed, it was Russell’s attempt to bring together Cantorian conceptions of set theory with practical measures of the actual world that begat the definitive paradox of set theory that bear Russell’s name, and the responses to which have in large measure shaped post-Cantorian mathematics. Russell gives the following account of his discovery of his eponymous paradox in his Autobiography:

Cantor had a proof that there is no greatest number, and it seemed to me that the number of all the things in the world ought to be the greatest possible. Accordingly, I examined his proof with some minuteness, and endeavoured to apply it to the class of all the things there are. This led me to consider those classes which are not members of themselves, and to ask whether the class of such classes is or is not a member of itself. I found that either answer implies its contradictory.

Bertrand Russell, The Autobiography of Bertrand Russell, Vol. II, 1872-1914, “Principia Mathematica”

None of the great problems of philosophical logic from this era — i.e., the fruitful period in which Russell and several colleagues created mathematical logic — were “solved”; rather, a consensus emerged among philosophers of logic, conventions were established, and, perhaps most importantly, Zermelo’s axiomatization of set theory became the preferred mathematical treatment of set theory, which allowed mathematicians to skirt the difficult issues in philosophical logic and to focus on the mathematics of set theory largely without logical distractions.

It is an irony of intellectual history that the next great revolution in mathematics to follow after set theory — which latter is, essentially, the mathematical theory of the infinite — was to be that of computer science, which constitutes the antithesis of set theory in so far as it is the strictest of strict finitisms. It would be fair to characterize the implicit theoretical position of computer science as a species of ultra-finitism, since computers cannot formulate even the most tepid potential infinite. All computing machines have an upper bound of calculation, and this is a physical instantiation of the theoretical position of ultra-finitism. This finitude follows from embodiment, which a computer shares with the world itself, and which therefore makes ultra-finite computing consistent with an ultra-finite world. In an ultra-finite world, it is possible that the finite may exhaust the finite and computational omniscience realized.

The universe defined by the Big Bang and all that followed from the Big Bang is a finite universe, and may in virtue of its finitude admit of exhaustive calculation, though this finite universe of observable cosmology may be set in an infinite context. Indeed, even the finite universe may not be as rigorously finite as we suppose, given that the limitations of our observations are not necessarily the limits of the real, but rather are defined by the limit of the speed of light. Leonard Susskind has rightly observed that what we observe of the universe is like being inside a room, the walls of which are the distant regions of the universe receding from us at superluminous velocity at the point at which they disappear from our view.

Recently in The Size of the World I quoted this passage from Leonard Susskind:

“In every direction that we look, galaxies are passing the point at which they are moving away from us faster than light can travel. Each of us is surrounded by a cosmic horizon — a sphere where things are receding with the speed of light — and no signal can reach us from beyond that horizon. When a star passes the point of no return, it is gone forever. Far out, at about fifteen billion light years, our cosmic horizon is swallowing galaxies, stars, and probably even life. It is as if we all live in our own private inside-out black hole.”

Leonard Susskind, The Black Hole War: My Battle with Stephen Hawking to make the World Safe for Quantum Mechanics, New York, Boston, and London: Little, Brown and Company, 2008, pp. 437-438

This observation has not yet been sufficiently appreciated (as I previously noted in The Size of the World). What lies beyond Susskind’s cosmic horizon is unobservable, just as anything that disappears beyond the event horizon of a black hole has become unobservable. We might term such empirical realities just beyond our grasp empirical unobservables. Empirical unobservables include (but are presumably not limited to — our “out” clause) all that which lies beyond the event horizon of Susskind’s inside-out black hole, that which lies beneath the event horizon of a black hole as conventionally conceived, and that which lies outside the lightcone defined by our present. There may be other empirical unobservables that follow from the structure of relativistic space. There are, moreover, many empirically inaccessible points of view, such as the interiors of stars, which cannot be observed for contingent reasons distinct from the impossibility of observing certain structures of the world hidden from us by the nature of spacetime structure.

What if the greater part of the universe passes in the oblivion of the empirical unobservables? This is a question that was posed by a paper appeared that in 2007, The Return of a Static Universe and the End of Cosmology, which garnered some attention because of its quasi-apocalyptic claim of the “end of cosmology” (which sounds a lot like Heidegger’s proclamation of the “end of philosophy” or any number of other proclamations of the “end of x“). This paper was eventually published in Scientific American as The End of Cosmology? An accelerating universe wipes out traces of its own origins by Lawrence M. Krauss and Robert J. Scherrer.

In calling the “end of cosmology” a “quasi-apocalyptic” claim I don’t mean to criticize or ridicule the paper or its argument, which is of the greatest interest. As in the subtitle of the Scientific American article, it appears to be the case that an accelerating universe wipes out traces of its own origins. If a quasi-apocalyptic claim can be scientifically justified, it is a legitimate and deserves our intellectual respect. Indeed, the study of existential risk could be considered a scientific study of apocalyptic claims, and I regard this as an undertaking of the first importance. We need to think seriously about existential risks in order to mitigate them rationally to the extent possible.

In my posts on the prediction and retrodiction walls (The Retrodiction Wall and Addendum on the Retrodiction Wall) I introduced the idea of effective history, which is that span of time which lies between the retrodiction wall in the past and the prediction wall in the future. One might similarly define effective cosmology as consisting of that region or those regions of space within the practical limits of observational cosmology, and excluding those regions of space that cannot be observed — not merely what is hidden from us by contingent circumstances, but that which are are incapable of observing because of the very structure of the universe and our place (ontologically speaking) within it.

There are limits to what we can know that are intrinsic to what we might call the human condition, except that this formulation is anthropocentric. The epistemic limits represented by effective history and effective cosmology are limitations that would hold for any sentient, conscious organism emergent from natural history, i.e., would hold for any peer species. Some of these limitations are limitations intrinsic to our biology and to the kind of mind that is emergent from biological organisms. Some of these limitations are limitations intrinsic to the world in which we find ourselves, and the vantage point from within the cosmos that we view our world. Ultimately, these limitations are one and the same, as the kind of biological beings that we are is a function of the kind of cosmos in which we have emerged, and which has served as the context of our natural history.

Within the domains of effective history and effective cosmology, we are limited further still by the non-quantifiable aspects of the world noted above. Setting aside non-quantifiable aspects of the world, what I have elsewhere called intrinsically arithmetical realities are a paradigm case of what remains computable once we have separated out the non-computable exceptions. (Beyond the domains of effective history and effective cosmology, hence beyond the domain of computational omniscience, there lies the infinite context of our finite world, about which we will say no more at present.) Intrinsically arithmetical realities are intrinsically amenable to quantitative methods are potentially exhaustible by computational omniscience.

Some have argued that the whole of the universe is intrinsically arithmetical in the sense of being essentially mathematical, as in the “Mathematical Universe Hypothesis” of Max Tegmark. Tegmark writes:

“[The Mathematical Universe Hypothesis] explains the utility of mathematics for describing the physical world as a natural consequence of the fact that the latter is a mathematical structure, and we are simply uncovering this bit by bit.”

The Mathematical Universe by Max Tegmark

Tegmark also explicitly formulates two companion principles:

External Reality Hypothesis (ERH): There exists an external physical reality completely independent of us humans.

…and…

Mathematical Universe Hypothesis (MUH): Our external physical reality is a mathematical structure.

I find these formulations to be philosophically naïve in the extreme, but as a contemporary example of a perennial tradition of philosophical thought Tegmark is worth citing. Tegmark is seeking an explicit answer to Wigner’s famous question about the “unreasonable effectiveness of mathematics.” It is to be expected that some responses to Wigner will take the form that Tegmark represents, but even if our universe is a mathematical structure, we do not yet know how much of that mathematical structure is computable and how much of that mathematical structure is not computable.

In my Centauri Dreams post on SETI, METI, and Existential Risk I mentioned that I found myself unable to identify with either the proponents of unregulated METI or those who argue for the regulation of METI efforts, since I disagreed with key postulates on both sides of the argument. METI advocates typically hold that interstellar flight is impossible, therefore METI can pose no risk. Advocates of METI regulation typically hold that unintentional EM spectrum leakage is not detectable at interstellar distances, therefore METI poses a risk we do not face at present. Since I hold that interstellar flight is possible, and that unintentional EM spectrum radiation is (or will be) detectable, I can’t comfortably align myself with either party in the discussion.

I find myself similarly hamstrung on the horns of a dilemma when it comes to computability, the cosmos, and determinism. Computer scientists and singulatarian enthusiasts of exponential increasing computer power ultimately culminating in an intelligence explosion seem content to assume that the universe is not only computable, and presents no fundamental barriers to computation, but foresee a day when matter itself is transformed into computronium and the whole universe becomes a grand computer. Critics of such enthusiasts often take the form of denying the possibility of AI or denying the possibility of machine consciousness, denying this or that is technically possible, and so on. It seems clear to me that only a portion of the world will ever be computable, but that portion is considerable and that a great many technological developments will fundamentally change our relationship to the world. But no matter how much either human beings or machines are transformed by the continuing development of industrial-technological civilization, non-computable functions will remain non-computable. This I cannot count myself either as a singulatarian or a Luddite.

How are we to understand the limitations to computational omniscience imposed by the limits of computation? The transcomputational problem, rather than laying bare human limitations, points to the way in which minds are not subject to computational limits. Minds as minds do not function computationally, so the evolution of mind (which drives the evolution of civilization) embodies different bounds and different limits than the Bekenstein bound and Bremermann’s limit, as well as different possibilities and different opportunities. The evolutionary possibilities of the mind are radically distinct from the evolutionary possibilities of bodies subject to computational limits, even though minds are dependent upon the bodies in which they are embodied.

Bremermann’s limit is 1093, which is somewhat arbitrary, but whether we draw the line here or elsewhere it doesn’t really matter for the principle at stake. Embodied computing must run into intrinsic limits, e.g., from relativity — a computer that exceeded Bremerman’s limit by too much would be subject to relativistic effects that would mean that gains in size would reach a point of diminishing returns. Recent brain research was suggested that the human brain is already close to the biological limit for effective signal transmission within and between the various parts of the brain, so that a larger brain would not necessarily be smarter or faster or more efficient. Indeed, it has been pointed out the elephant and whale brains are larger than mammal brains, although the encephalization quotient is much higher in human beings despite the difference in absolute brain size.

The function of organic bodies easily peaks over 1093. The Wikipedia entry on the transcomputational problem says:

“The retina contains about a million light-sensitive cells. Even if there were only two possible states for each cell (say, an active state and an inactive state) the processing of the retina as a whole requires processing of more than 10 to the 300,000 bits of information. This is far beyond Bremermann’s limit.”

This is just the eye alone. The body has far more nerve ending inputs than just those of the eye, and essentially a limitless number of outputs. So exhausting the possible computational states of even a relatively simple organism easily surpasses Bremermann’s limit and is therefore transcomputational. Some very simple organisms might not be transcomputational, given certain quantifiable parameters, but I think most complex life, and certainly things are complex as mammals, are radically transcomputational. Therefore the mind (whatever it is) is embodied in a transcomputational body, of which no computer could exhaustively calculate its possible states. The brain itself is radically transcomputational with its 100 billion neurons (each of which can take at minimum two distinct states, and possibly more).

Yet even machine embodiments can be computationally intractable (in the same way that organic bodies are computationally intractable), exceeding the possibility of exhaustively calculating every possible material state of the mechanism (on a molecular or atomic level). Thus the emergence of machine consciousness would also supervene upon a transcomputational embodiment. It is, at present, impossible to say whether a machine embodiment of consciousness would be a limitation upon that consciousness (because the embodiment is likely to be less radically transcomputational than the brain) or a facilitation of consciousness (because machines can be arbitrarily scaled up in a way that organic bodies cannot be).

Since the mind stands outside the possibilities of embodied computation, if machine consciousness emerges, machine embodiments will be as non-transparent to machine minds as organic embodiment is non-transparent to organic minds, but the machine minds, non-transparent to their embodiment as they are, will have access to energy sources far beyond any resources an organic body could provide. Such machine consciousness would not be bound by brute force calculation or linear models (as organic minds are not so bound), but would have far greater resources at its command for the development of its consciousness.

Since the body that today embodies mind already far exceeds Bremermann’s limit, and no machine as machine is likely to exceed this limit, machine consciousness emergent from computationally tractable bodies may, rather than being super-intelligent in ways that biologically derived minds can never be, may on the contrary be a pale shadow of an organic mind in an essentially transcomputational body. This gives a whole new twist to the much-discussed idea of the mind’s embodiment.

Computation is not the be-all and end-all of of mind; it is, in fact, only peripheral to mind as mind. If we had to rely upon calculation to make it through our day, we wouldn’t be able to get out of bed in the morning; most of the world is simply too complex to calculate. But we have a “work around” — consciousness. Marginalized as the “hard problem” in the philosophy of mind, or simply neglected in scientific studies, consciousness enables us to cut the Gordian Knot of transcomputability and to act in a complex world that far exceeds our ability to calculate.

Neither is consciousness the be-all and end-all of mind, although the rise of computer science and the increasing role of computers in our life has led many to conclude that computation is primary and that it is consciousness is that is peripheral. and, to be sure, in some contexts, consciousness is peripheral. In many of the same contexts of our EEA in which calculation is impossible due to complexity, consciousness is also irrelevant because we respond by an instinct that is deeper than and other than consciousness. In such cases, the mechanism of instinct takes over, but this is a biologically specific mechanism, evolved to serve the purpose of differential survival and reproduction; it would be difficult to re-purpose a biologically specific mechanism for any kind of abstract computing task, and not particularly helpful either.

Consciousness is not the be-all and end-all not only because instinct largely circumvents it, but also because machines have a “work around” for consciousness just as consciousness is a “work around” for the limits of computability; mechanism is a “work around” for the inefficiencies of consciousness. Machine mechanisms can perform precisely those tasks that so tax organic minds as to be virtually unsolvable, in a way that is perfectly parallel to the conscious mind’s ability to perform tasks that machines cannot yet even approach — not because machines can’t do the calculations, but because machines don’t possess the “work around” ability of consciousness.

It is when computers have the “work around” capacity that conscious beings have that they will be in a position to effect an intelligence explosion. That is to say, machine consciousness is crucial to AI that is able to perform in that way that AI is expected to perform, though AI researchers tend to be dismissive of consciousness. If the proof of the pudding is in the eating, well, then it is consciousness that allows us to “chunk the proofs” (i.e., to divide the proof into individually manageable pieces) and get to the eating all the more efficiently.

. . . . .

signature

. . . . .

Grand Strategy Annex

. . . . .

Saturday


There were a number of old cars in the town, so I took this one street scene with three cars from the 1930s. There are also a couple of contemporary cars, as well as a few other clues that this is not an old photograph, but if I had been a bit more careful I see that it would have been possible to take a picture that looks like a vintage post card.

There were a number of old cars in the town, so I took this one street scene with three cars from the 1930s. There are also a couple of contemporary cars, as well as a few other clues that this is not an old photograph, but if I had been a bit more careful I see that it would have been possible to take a picture that looks like a vintage post card.

The Historic Quarter of the City of Colonia del Sacramento is on the UNESCO World Heritage list, which is always a good indication that a particular place will be worth a stop and a look. I’ve been to several UNESCO sites on the World Heritage List, and I always enjoy them, but it almost always is the case that they are different from what I expected. Colonia was not exactly what I expected, but that was great. That is why one travels: in order to have one’s concepts corrected by one’s percepts, just as one thinks in order to have one’s percepts corrected by one’s concepts. Thought and experience are an indissoluble unity; it is when we divide them and compartmentalize them that we get into trouble.

Colonia 4

Why travel? What is to be gained from travel? Travel is all about challenging assumptions. One could simply stay at home, read books, look at travel brochures, and watch travel videos, convincing oneself that one had learned all there is to know about a place, and never bother to go there oneself. But we all know that you do need to eventually go to the place — whatever the place happens to be — if you want to understand it on its own terms, rather than attempting to understand a place one has never visited on the basis of one’s preconceived idea of the place.

Colonia 1

To say that thorough knowledge of a place is not adequate to saying that one really knows a place made me realize that this suggests a generalization of a thought experiment in the philosophy of mind known as “Mary’s Room.” In the Mary’s Room thought experiment, Mary is a scientist locked in a black and white room, who studies everything that there is to know about color vision. After perfecting her knowledge of color, she leaves her black and white room, and suddenly experiences what it is like to actually see color. The question, from a philosopher’s point of view, is this: when Mary leaves her room, does she learn anything? This thought experiment is also known as the “knowledge argument,” in so far as it points to knowledge that can be attained only through conscious experience.

Colonia 3

Putting Mary’s room and the knowledge argument in the context of travel suggests a generalization of the knowledge argument: suppose, in isolation of the object of knowledge studied, one learns all that there is to learn about a given object of knowledge. Say that one learns all that there is to learn about Colonia del Sacramento. After learning about Colonia del Sacramento, does one learn anything by traveling there? Even the most experienced of travelers know that you learn something by visiting a place that you cannot learn by all the research you might possibly conduct. Another way to put this would be to say that there is something that it is like to be in a place — a formulation parallel to Nagel’s famous formulation about there being something that it is like to be a bat.

Colonia 2

The conscious experience of a place is a source of knowledge not attainable through study. As I write this I realize that this argument entails that such knowledge is ineffable, otherwise, someone who visited a place and realized what was lacking in its description could simply write it down after having visited, and every subsequent visitor would thereafter visit the place with no surprise at all, and no new knowledge would be attained by such a visit. And yet we know it isn’t like that.

. . . . .

signature

. . . . .

Grand Strategy Annex

. . . . .

%d bloggers like this: