Latest / Disintegrator / 17. Computation is Computation (w/ M. Beatrice Fazi)

Transcript

17. Computation is Computation (w/ M. Beatrice Fazi) · Disintegrator

• 0:00
  Disintegrator Podcast
• 0:07
  This episode is kind of like a little birthday present to myself, especially in that it includes
• 0:13
  our number one most requested guest and one of our most giddily anticipated guests
• 0:20
  M. Beatrice Fazi. Fazi is a philosopher working in philosophy of computation, philosophy
• 0:28
  of technology, media philosophy. She's a reader in digital humanities in the school of
• 0:33
  media, arts and humanities at the University of Sussex, where she's currently also the head
• 0:37
  of the Department of Media and Cultural Studies. Fazi's PhD thesis and book, contingent
• 0:44
  computation, abstraction, experience and indetermacy in computational aesthetics published way
• 0:51
  back in 2018, like totally rings fresher than ever today. Seriously, like this book is
• 0:56
  so unbelievably important to the present moment where philosophy has this resurgent power
• 1:01
  and capacity around the world as new developments of technology and their covalent emeshment
• 1:07
  in society and politics force us to ask seriously major, seriously basic, seriously like fundamental
• 1:14
  questions about what these things are and what they are doing and what they mean to us.
• 1:20
  We spend a good deal of this episode, honestly like 90% of this episode, talking through
• 1:26
  some definitions and we do this for one reason. Like Fazi is like one of the few maybe like
• 1:31
  only people who actually provides us with like real accessible definitions for things like
• 1:37
  computation, algorithm, abstraction, thought and talking through those definitions is like
• 1:43
  really grounding and really really really helpful. And yeah, you can probably imagine like
• 1:49
  based on all of the above, like the episode is pretty dense, like it's pretty dense, but
• 1:56
  it's honestly like actually super unreal once you start to realize what Fazi is telling
• 2:01
  us — that computation is a distinct and autonomous form of thought whose distinctness and autonomy
• 2:08
  are incredibly important to understand right now at this moment. There are some major
• 2:15
  deft blows to a lot of, you know, mainstream media theory, to some 80s and beyond
• 2:21
  philosophical speculation, and implicitly to political orientations toward computation
• 2:27
  that are packed within here. So buckle up. Totally, literally top five lifetime hero
• 2:33
  of the pod M. Beatrice Fazi. FAZI: What is computation? Okay, historically obviously there are
• 2:42
  multiple accounts of computation of them these accounts clash with one another because they
• 2:48
  originate in different fields, or aim to respond to different agendas in different goals.
• 2:55
  There is a text quite like by the philosopher Brian Cantwell Smith, an essay called “The Foundations
• 3:02
  of Computing" which came out now 20 years ago in 2003. And the essay really maps some
• 3:07
  of the construals shaping our technological, cultural, and scientific understandings of computation.
• 3:16
  The text, which I was reading just recently, can be a useful entry point for your audience
• 3:22
  to have a sense of the intellectual debate on and around computation, although I should note
• 3:28
  that in the same text Brian also says that ultimately computation cannot be defined.
• 3:35
  And I do disagree with that statement. I think it's very important to ask the "what is it?” question —
• 3:42
  what Socrates will call the “ti esti” question. And, also, I believe in the importance of defining
• 3:49
  words and terms especially within philosophy. So, in my view, philosophy must define, not just
• 3:54
  describe, it should demarcate problems... yes make them more circumscribed. And at the same time
• 4:00
  open up these problems to generality of a definition. So, let’s start from definitions, shall we?
• 4:08
  My own working definition of computation is one I have formulated years ago, but I still
• 4:15
  stand by it. I understand computation as a method, a method of organizing, systematizing,
• 4:23
  arranging reality. A method that is quite peculiar and also distinctive because the organization
• 4:30
  of the real that it allows for is carried out by a logical, discrete, finite and quantitative
• 4:36
  means. Much of my work in philosophy of technology is really focusing on understanding
• 4:41
  about the ontological and epistemological specificity of the computational method. For me, ontology
• 4:47
  and epistemology always go together. There is a phrase that I wrote a year ago. I said
• 4:54
  something like "computation is computation." And that has been picked up by others to describe
• 5:00
  my work and I'm fine with that because that phrase is meant, although it's a tautology
• 5:04
  and obviously within the limits of a tautology, it's meant precisely to summarize one of the
• 5:09
  key goals of my philosophical research, which is to approach computation on its radically
• 5:15
  singular terms. So for instance computation as a method, it's a method, we said, and as a
• 5:22
  method, is strongly related to mathematics. To compute is to a large extent to calculate.
• 5:28
  And yet computation is not only mathematics, because computation has a distinctively
• 5:33
  processual character. So a computational procedure has to be run. Insofar as a computational
• 5:40
  procedure is a process, it's got a different ontoepistemic status from a mathematical formula
• 5:47
  for instance. So this process that computation always is has to bring itself into existence.
• 5:53
  And the only way as to do it is by producing itself alongside its own world. And this striving
• 6:00
  for actualization is really something characteristic of computation, which we said, according
• 6:07
  to my own definition of computation is a method of systematizing and organizing. Yes, but
• 6:12
  those systematizations and those organizations are also constructions. Computation is a method
• 6:19
  of and for construction. Computation always constructs, in my view. Okay, so we said computation, a
• 6:27
  computational procedure, stands alongside a mathematical formula as a distinct entity,
• 6:32
  but I'd like also to add, that a computational procedure stands alongside a living organism
• 6:38
  and a physical object as an equally distinct entity. And I've written a lot about the differences
• 6:44
  between ontologies of life, on the one hand, and ontology of mechanisms, on the other. I'm
• 6:50
  referring to my work on computational automation and my critique of the simulative paradigm in
• 6:56
  artificial intelligence research. When I stress the difference between a computational procedure
• 7:02
  and a living organism, obviously I'm implicitly criticizing pan-computationalist views.
• 7:09
  I don't think that nature computes. I don't believe the universe is a giant Turing machine,
• 7:14
  for instance, but with that statement I also want to kind of underline the inadequacy of any vitalist
• 7:22
  approach to computation, which in my view do not consider the ontopistemic specificity of
• 7:28
  computational processes. Okay, so computation is different from a mathematical formula,
• 7:34
  it's not a living organism, and then perhaps the most contentious one. It's also we need to stress
• 7:40
  the differences between a computational procedure and a physical object. By saying that,
• 7:46
  obviously, I don't want to restate any mind-body dualis,, by all means, I'm not doing that.
• 7:52
  And obviously it's correct to understand computation as a method that it's always realized
• 7:58
  to logic and through matter, through software and through hardware. In the past, maybe two decades, fifteen years or so,
• 8:05
  digital media studies, critical theory of technology, science and technology studies,
• 8:11
  philosophy of computation, they all emphasized the materiality of computation,
• 8:16
  of computing practices. And this is very interesting because as happened in parallel,
• 8:21
  with different emerging paradigms in cognitive science, which also stressed
• 8:27
  the embodiment and the situatedness of computational procedures and practices,
• 8:33
  and often they do that through engagements with phenomenological critiques of formalization.
• 8:38
  And obviously I appreciate the work, and I also build on the work, and I do not deny the importance
• 8:47
  of considering the materiality of computation, but at the same time I think that stressing too much
• 8:53
  materiality of computing practices risks us to make us forget that the computational method
• 9:00
  would not really amount to much or account for much if it wasn't for a relationship with a
• 9:05
  dimension of intelligibility, a relation that is established at the operational level
• 9:12
  of every single algorithmic procedure, and in fact it's inscribed within the very definition
• 9:17
  of what an algorithmic procedure is. So, all intelligibility must reckon and contend
• 9:24
  with its embedding. However, computation is powerful and it's pervasive and makes worlds,
• 9:30
  constructs, as we said, precisely because it has a relation with what is apprehensible only through
• 9:36
  forms of abstractive activity. And that's what I mean by intelligibility, what is apprehensible only through
• 9:42
  forms of abstractive activity. Without the relation, quite frankly, a silicon
• 9:48
  chip is just sand. So a central claim of my work is the computational system offer
• 9:56
  a different order of intelligibility, and if we are to understand computation's operations
• 10:02
  of structuring, we must address the order of intelligibility in its own terms.
• 10:09
  Finally, to conclude, this very long answer to this question, these are two direct implications
• 10:14
  for the kind of speculative philosophy of computation I'm interested in pursuing. First of all,
• 10:20
  speculative philosophy of computation needs to acknowledge the specific contributions
• 10:25
  of computational formalization, and secondly, it should not include these contributions
• 10:31
  by conflating computation with related formal sciences or with its employment,
• 10:36
  in cultural and social practices. To sum up again, computation is computation.
• 10:42
  What a great way to kick off the episode. I mean an unbelievably killer definition of computation,
• 10:48
  both positively and negatively. Following that up a little bit, like, where does computation
• 10:54
  come from? Is computation anthropogenic? Is computation something that is discovered or something
• 11:02
  that is constructed? Oh, that's a brilliant question, and it relates obviously, of course,
• 11:08
  to the question, a similar question in maths: are numbers discovered or constructed?
• 11:14
  I'm tempted to say that it's constructed in so far as everything is really
• 11:20
  is a construction in my own kind of ontology. So it's not constructed because it can only live within
• 11:29
  like the social, cultural domain. I'm very much interested also in trying to explore whether
• 11:35
  the symbolic order, to use an old Lacanian term, can exist really beyond the cultural.
• 11:41
  So in a way, yeah, it's constructed because it has to be made. I'm giving you a
• 11:50
  tentative question here. So yeah, I don't have the final question, but I will be tempted to say that
• 11:54
  it's constructed in so far it needs to construct itself alongside its own world. So it might be just
• 12:00
  constructed by its own activity. It's not a given. It's not something that we just go outside and
• 12:07
  we find computation, I don't know, the ants computing or the solar system computing. No, I don't believe
• 12:14
  that it's there are nice metaphors that we use. Cultural metaphors, so like, nature
• 12:19
  computes, so on and so forth… But it's constructed insofar as it's always mediated by the symbolic.
• 12:27
  But yeah, great question. Okay, constructed, but it's not like it's not like subordinated to or
• 12:35
  thoroughly downstream of social construction, right? Like I feel like that's a point that I think
• 12:39
  a lot of people, yeah, I think I'm just trying to underline it because that’s the point that's kind of
• 12:44
  endlessly talked about at the moment is the sort of social construction of all of this stuff from,
• 12:49
  you know, computation to AI. No, I would say I'm a constructivist but not a social constructivist.
• 12:56
  So it's more like an ontological constructivism of the relationship between what can be thought and
• 13:02
  what is instead found in nature. Som this relation between thought and world is always in itself a
• 13:09
  construction because it's mediated. It’s not just given, but it's definitely not a social
• 13:14
  constructivism. And in a sense, more than one sense, in fact, in my work, I always stress the autonomy
• 13:21
  of computation. So in classic or traditional kind of media theory, you know, the two,
• 13:27
  like the two camps, one side McLuhan and with the media as extension of men and the other side,
• 13:35
  Kittler as the autonomy of media. I would definitely side with Kittler, although I'm not Kittlerian,
• 13:41
  there is, there are problems with Kittler, in my view, but definitely one very, very good thing
• 13:49
  that I take from Kittler is the autonomy of the technological. I believe in a technological
• 13:54
  a priori, in a sense. Yeah. Great, that was a really comprehensive intro to computation. I guess
• 14:01
  the next question, logical question is to discuss contingency, which plays a crucial role in your
• 14:09
  research and the role that it has within computation. And I guess the broad question of what is
• 14:17
  contingent computation... how would you, how would you define it? So the book came out in 2018. And since then
• 14:25
  we are more and more talking about contingency in relation to technological systems. Different people
• 14:31
  are trying to demonstrate the technology, in fact, does involve contingency in one way or another
• 14:37
  and I'm glad this is happening, although I always need, to in a way, stress the very specific way
• 14:43
  in which I understand contingency in computation. Mine was a quite specific proposition
• 14:49
  with a lot of caveats. So let's start from what is not. Contingent computation is not unpredictability,
• 14:57
  it's not uncertainty, it's not something that might happen or might not happen, it's not the improbable,
• 15:04
  this is not the kind of contingency that I find in computation. It's not the contingency of
• 15:11
  empirical chance, it's not an accident, it's not randomness either, insofar as randomness
• 15:18
  always implied a kind of connection to probability. And then most certainly, it's not glitch -
• 15:25
  this is something I always need to specify when giving talks in art context, this is not the contingency
• 15:31
  of a machine running amok, but it's the contingency of a machine doing precisely what is supposed to do.
• 15:38
  Artists love talking about glitch. And it's a phenomenon profoundly interesting glitch, but glitch is
• 15:46
  in the eye of the beholder, I always say because to an extent it doesn't exist in the machine.
• 15:50
  The kind of glitch, at least, in which that it's expressed artistically in many many
• 15:55
  profoundly interesting instances. But contingent computation is not glitch, definitely.
• 16:01
  The key point of my argument about contingent computation is that this is a kind of
• 16:06
  indeterminacy which is profoundly different from the indetermination of life,
• 16:11
  of living systems, of lived experience. But it's in fact an indeterminacy that's formal,
• 16:18
  mechanical, and always internal to computational processing and which it’s, most importantly,
• 16:23
  inscribed into the logical and mathematical definition of what an algorithmic procedure is.
• 16:29
  So in this respect the book “Contingent Computation” pursues like a kind of divorce between
• 16:36
  the concept of the contingent from the empirical. Obviously, I recognize and acknowledge
• 16:44
  computation might be modeled upon the mutability of the real word, for lack of a better word.
• 16:51
  And through interactive, embedded operations might become more powerful and adaptable. This is really what
• 16:58
  much of current machine learning is doing. And obviously, algorithms might be enmeshed with the
• 17:05
  empirical mutability of life and living beings and lived experience through different forms of
• 17:11
  applications, implementations and performance. However, the very, very key point of my argument about
• 17:17
  contingent computation is that computation is already contingent before any implication on association
• 17:25
  with what I call the associated milieu, which is like really computation plus life, or computation plus
• 17:32
  culture, computation plus society, economy and so on and so forth. That was the negative answer
• 17:37
  also in the sense why this is not contingent computation. So why this contingent computation?
• 17:43
  It’s, quite simply, it should be viewed as the maximally indeterminate. It's the indeterminacy that I find
• 17:51
  by engaging with certain histories and trajectories in the history of computability theory,
• 17:56
  is the indeterminacy that I find at the very heart of computing itself.
• 18:01
  It’s an indeterminacy that is central to the formal operations of computation, again, before we
• 18:07
  might consider any external source of indetermination. I think contingency brings us also to the history of
• 18:16
  computational theory. There are two key ideas that you discuss. One is Gödel's idea or discussion of
• 18:22
  incompleteness and also Turing’s incomputibility. And the question there, which is I think it's really
• 18:28
  interesting, you know, tracing the idea of contingency back to some of these foundational texts.
• 18:34
  What is the connection? How do you trace this connection with these two key nodes?
• 18:39
  I could talk about Gödel and Turing for hours, but you'll be spared,
• 18:44
  I have also a short version of this. So, as we were trying to express, one of the central and
• 18:52
  hopefully more innovative theses of the work that developed for the book “Contingent Computation"
• 18:57
  is the argument that in order to find novelty in computation which should look no further than
• 19:02
  the axiomatic method of computation itself. I'll just briefly explain what an axiom is. An axiom is
• 19:09
  self-evident truth. It's the postulate from which consequences are inferentially derived.
• 19:16
  Computational systems are formal axiomatic systems. There is no escaping from that. There's really no
• 19:23
  avoiding of that. And because they are axiomatic, they too implicate a starting point that is
• 19:29
  self-evident and which defines the inferential steps that can be taken from it toward a result.
• 19:35
  From this point of view, computer programs provide an output that is already implicit in the
• 19:43
  input. And you see if you want to find a kind of a form of extreme necessity in computation, that
• 19:49
  would be precisely this axiomatic nature, that everything is already kind of present and predetermined.
• 19:54
  In the work I did, I proposed not to discard the axiomatics of computation but in fact to
• 20:04
  reengage with it. And I proposed what is the factor of reconceptualization of it. I took some negative
• 20:12
  results and I turned them into positive results. So Gödel and Turing, in a way, bought found limits
• 20:19
  to computation, and for me those limits became really the potential inherent to computation itself.
• 20:27
  Something that at times my you know really sound a bit weird but it's quite justified in my view.
• 20:38
  So briefly what happened in the 30s? In 1931 the logician Kurt Gödel determined the incompleteness
• 20:47
  of formal axiomatic systems by demonstrating that there are propositions that appear to be true
• 20:54
  when viewed from the outside of the system, but which cannot be proved or disproved
• 21:00
  from within the system itself. This was 1931-31-32. Then in 1936 the mathematician
• 21:08
  Alan Turing, who was working in England at the time at Cambridge, he drew on Gödel's discovery
• 21:16
  of incompleteness to demonstrate that some functions are incomputable. So he strikingly described
• 21:24
  the limits of computing machines before any such machine was built.
• 21:31
  Gödel addressed the theoretical consistency and completeness of formal rules for the description
• 21:38
  of logic and mathematics and Turing, on the other hand, investigated the possibility of finding a
• 21:44
  mechanical, finite step and formally expressed process that would set the standard for deductive
• 21:52
  procedures. What Turing did then, he formalized the computational method in terms of a procedure,
• 21:58
  an algorithm, that would solve a problem in a finite number of steps. So, computation in the sense, is
• 22:05
  formalized. I said before that my definition of computation is, computation is a method. Yes, it's a method,
• 22:11
  but through Turing, this foundational work done in the 1930s, not only by Turing, but other people
• 22:17
  also were working on similar formalization of computation, computation becomes also a theory.
• 22:24
  So it's a method but it's also a theory. Through that formalization of computation, Turing's,
• 22:30
  however, it was revealed the very challenging existence of incomputable problems,
• 22:37
  that is, problems that cannot be solved by algorithmic means because the steps
• 22:43
  involved in their processing are not finite, but infinite. So Gödel's and Turing's
• 22:50
  respective results have been traditionally read in terms of failure of formal,
• 22:56
  axiomatic systems. However, iny my own philosophical reading, incompleteness and incomputibility,
• 23:01
  as I said, are not errors, so they are not stopping points, but a mark of computation’s potential.
• 23:07
  They show that there is an indeterminacy that pertains not only to the empirical realm,
• 23:12
  but to the formal too. And in this sense, they show, they reveal computations very own contingency.
• 23:21
  As we were saying before, according to my definition of what contingency in computation is.
• 23:25
  Again, this is not the contingency of computation doing something that it's not supposed to do.
• 23:32
  This kind of contingency doesn't contradict the functionality of computation, and quite the opposite,
• 23:37
  it’s this kind of contingency in this relation within an inherend indeterminacy at the level of the
• 23:44
  logical constitution of the algorithm itself, that makes computation work. So something that Roberto and I
• 23:52
  are really interested in at the moment, something that we’re currently working on a book on, is the concept of scale.
• 23:57
  And we're taking scale from engineering, from design, from politics and economics. And we understand
• 24:03
  scale to be the kind of like interoperability of some object across or within certain domains of,
• 24:08
  let's say, size or complexity, magnitude, and something interesting to us. And something we talked about
• 24:15
  a bit with Georgina Voss is the kind of affordances of digital technology and specifically
• 24:21
  computation with respect to scale. So I'm wondering like, does computation scale? How do you understand
• 24:29
  the concept of scale within computation? And does computation as an object or as a method change its
• 24:36
  nature at different scales? In a way, scale doesn't fit feature in my work as a concept per se.
• 24:45
  But if I am here to think with you about scale and how scale can be engaged alongside the things
• 24:49
  I'm interested in addressing and I've been studying for years, I would say that for me the most
• 24:56
  multi-scalar or scalable… or at the center of scale itself is the process, the concept or whatever we
• 25:05
  want to call it, of abstraction, which is really the core of my work. I always say, jokingly that yes,
• 25:13
  I studied technology and machines, but really what I study is abstraction, with an A, Abstraction.
• 25:20
  Yeah, recognizing that computation is abstract and operates through abstraction is one of the most
• 25:27
  concrete things we can do. And abstraction might as well be the most multi-scalar of all phenomena,
• 25:35
  although not a phenomenon per se. It's something outside of computation because not only computing
• 25:41
  abstracts, we do abstracts, but like there are different forms and modes of abstraction,
• 25:47
  but it's also the concept that has got the largest impact in delimiting the scope of the field
• 25:56
  of philosophy of computation and of technology in large. So at the most like,
• 26:01
  the most molar or most molecular level, it's really abstraction, what would be a way for me to understand
• 26:11
  scale. What's the difference between abstraction and like generalizability? Yeah, yeah, that's another
• 26:19
  good question. It's the systematizing, striving in a way for me. abstraction... So at the moment,
• 26:28
  I'm triangulating three concepts for my new book. And I don't have the answer, the full answer yet...
• 26:34
  but I will at some point. This triangulation is between abstraction, representation, and thought.
• 26:41
  So, abstraction always has to represent. It’s a form of representation and vice versa representation
• 26:50
  is through abstraction, and thought also abstracts and represents. So, something is abstract - it's not just
• 26:58
  general, something universal. It's not just a matter of the universal of the particular,
• 27:04
  as the Medievals would say, because abstraction implies an activity of drawing away and also
• 27:11
  constructing and systematizing those abstractions alongside other abstractions. So there's always a
• 27:18
  constructivist element in abstraction as well. Yeah, and that's also going to speak on the concept of
• 27:27
  scale in relation to multiple planes of realizability. I think this might bring us back to
• 27:35
  another key concept that maybe we can discuss a little bit, which is that of the algorithm
• 27:40
  that you use in a specific way and maybe that's something we can discuss to some extent,
• 27:48
  the idea of the relationship there, which is almost performative between writing, doing
• 27:54
  transformative power of algorithm and so on. And yes, what is an algorithm? How do you envision
• 28:00
  an algorithm? Yeah, an algorithm is a step-by-step procedure to achieve an output from an input,
• 28:08
  but what is really interesting for me about algorithmic procedures, which, however are not
• 28:16
  just computational, like they can be instantiated in computational systems, but it's a category
• 28:22
  very much larger than that of computation in a sense… it's the fact that this step-by-step procedure
• 28:30
  is an act of discretization. So, it's breaking up something into a sequence of finite steps,
• 28:40
  whether it’s, I don't know, an act such as, I don't know, walking from here to there,
• 28:47
  or something like a perceived continuity that has to broken down into a sequence of finite steps.
• 28:55
  And the steps have to be finite precisely because if they are not, they wouldn't
• 28:58
  classify according to Turing's indefinitions of computability. And I'm interested in the algorithm as
• 29:08
  a form of discretization because, at the same time, to go back to what we were saying about abstraction,
• 29:15
  to abstract is also always to discretize. You have to, quite literally, if you take from the
• 29:21
  Latin, to draw away from the immanence of things. So, it might not be a transcendence per se,
• 29:29
  but it's definitely like a transcendental act of discretization. And again, according to
• 29:35
  Turing's model of computation, to compute, again, is to follow an effective procedure, in other words,
• 29:40
  an algorithm to solve a problem in a finite number of steps. In my study of algorithmic computation,
• 29:47
  which is really a study of abstractive discretization, I'm not interested in... I don't want to offer an
• 29:54
  apology for the, like, regimes of instrumental rationality, but I want to rework the manner in which
• 30:01
  it's possible to understand this discretization, which in any case, I think we do anyway. I don't
• 30:08
  think it's possible to exist outside abstraction and representation really. So it's a
• 30:15
  form of discretization, an act of discretizing or cutting away, separating from the buzzing and
• 30:23
  blooming confusion of just, you know, immanent life as this blob of sensation. It's something that
• 30:29
  is an inherent in every act of experience. For me, abstraction, it's now how you leave the world,
• 30:39
  but in fact, how you enter the world and you inhabit and you construct it. So, it's not something that
• 30:45
  reduces or detatches us from life, but it's how we enter life really. That's why I worked a lot
• 30:54
  on the relationship between abstraction and experience precisely to understand how this rapport is played
• 31:01
  out in computation, obviously, but also more generally. And the algorithm, to go back to your question
• 31:07
  Roberto, I mean, I'm interested mostly and primarily in algorithmic computation, precisely because I'm
• 31:14
  interested in understanding discretization, what this act of discretization is, and its rapport
• 31:20
  between finitude and infinitude. So to compound the complexity of all this further, like, how does algorithm
• 31:27
  or its kind of, you know, transduction here into the concept of discretization, how does that sit
• 31:32
  within the trinity of abstraction, representation and thought? I'm assuming it's clinging to the former
• 31:38
  and kind of flirting with the other two categories, like... does it touch the third category of thought?
• 31:43
  Okay, this is still like work in progress. In the kind of ontology I'm building,
• 31:49
  it's really like, ultimately, discretization might be the least common denominator between the all.
• 31:55
  I mean, is it the fact that … to abstract is to discretize, but isn't to represent to discretize too?
• 32:04
  And isn't thinking in fact an act of discretization? Whenever we think ... I mean, like the other day a
• 32:12
  colleague of mine said, "oh, I'm such an analog thinker." And I smiled, because for me, it's like a kind of
• 32:18
  contradiction here, because all things… I'm going to, yeah, I'm writing a piece on this, which is
• 32:23
  going to come out at some point about like, all thought is really digital, because all thought
• 32:27
  implies really like a critique of immanence, in a way. There’s not anything as such as thought
• 32:34
  that it's utterly eminented to sensation. There's always a kind of drawing away from sensation.
• 32:40
  And I've wrote a lot on Deleuze. I have lots of, like, philosophical debt toward Deleuze,
• 32:45
  in the sense that he informed a lot the way in which I think about many things in life, in philosophy,
• 32:52
  but the thing I really couldn't get along with Deleuze is basically his flattening down
• 32:57
  of thought onto sensation. And that's primarily in my opinion why he really doesn't get computers.
• 33:04
  Yes, so in my, in this triangle on a post-it, in these three words that I'm trying to work out,
• 33:13
  iscretization at the moment is the element that really unifies, for lack of a better word,
• 33:20
  the three concepts as the operation behind the three of them. So it's a plea also for the study
• 33:27
  of the digital in a kind of much more ontological sense. Another aspect of your thinking that I find,
• 33:35
  we find quite interesting, ... and I think in general it's really nice, the clarity, the conceptual
• 33:41
  clarity of how you frame things and how you define what you mean by things, which is sometimes lacking
• 33:47
  in general. And that's a bit of clarity, I think is fundamental. But to us it's very interesting how
• 33:54
  you frame the discussion of contingent computation within the aesthetic domain. And I guess the question
• 34:00
  there is how do you address the relationship between the aesthetic and the logical? What is the aesthetic?
• 34:04
  That are, like, two reasons why I frame the problem of contingent computation, of the issue of the
• 34:11
  question of contingency in computation, within an aesthetic domain. The aim of the book was to demonstrate
• 34:18
  the computation can be productive of novelty, ontological novelty, not artistic novelty,
• 34:25
  it can produce the new at an ontological level. And it's because I focus on the question of
• 34:30
  ontological novelty, the production of the new, I have to situate this speculation within an
• 34:37
  aesthetic domain. I add to start from aesthetics because aesthetics is indeed about the production of the new.
• 34:44
  And I draw from a lineage within continental philosophy, which understands aesthetics
• 34:50
  in ontological terms as a way to address reality itself at a metaphysical level. So aesthetics, in
• 34:55
  other words, is an openly metaphysical inquiry into the conditions of the real. In fact, into the
• 35:02
  potentiality of reality. This is a kind of pre-Hegalian reworking of the aesthetic inquiry. If we take
• 35:09
  Hegel as the moment in which, in the history of thought, when aesthetics became a theory of art,
• 35:15
  but also a reworking the while being pre-Hegalian, it's also post-Kantian because it proposes a
• 35:21
  study meant to address the production of reality in terms of its conditions of possibility.
• 35:28
  So this is kind of hardcore because there's a little bit of a rupture with aisthesis, right? Like in
• 35:34
  some... Yes, I was getting to that. Yeah, yeah. Okay, sorry. Okay, no, no, no, no. Yeah, no, tell me, why
• 35:41
  you saw it in that way? Because that's that's that's why I meant it. Well, I mean, because you're
• 35:45
  speaking about aesthetics in kind of ontological terms, right, as the production of new ontological forms.
• 35:51
  And it's therefore kind of existing on levels that are prior to the sensorium, or decoupled from
• 35:57
  the sensorium to some kind of meaningful extent. And of course, like, aisthesis is something that's very
• 36:01
  much kind of coupled to sensation. Yeah, yeah. So this sense of aesthetics as an
• 36:07
  ontological inquiry. So, an ontological aesthetics or an aesthetic ontology really comes from a lineage
• 36:15
  in in continental philosophy that really understood aesthetics as I said, not as a theory of art,
• 36:23
  but as a theory, the production of the new. Although that production of the new really comes from
• 36:30
  the original etymological sense of aesthetics as aisthesis, so Baumgarten, and of course
• 36:37
  aisthesis as knowledge that comes from the senses, and so on and so forth. What happened within
• 36:44
  the history of philosophy, like recent history of philosophy and history of philosophy
• 36:48
  as applied to media and computation and so on and so forth, then that production of the new
• 36:54
  … because in order to produce the new, you have to find a potentiality … Where does the new
• 37:00
  come from? You have to find, like, the potential. The potential, within aesthetic discourses, has been
• 37:07
  predicated mostly, and uniquely, on sensation. This is again Deleuze, but this is also the affective turn
• 37:15
  in media studies and cultural theory, and so on and so forth. So, when I said to you that there are two reasons
• 37:21
  why I had to situate contingent computation within aesthetics, reason number one was because,
• 37:26
  yes, it's about producing the new. So, if I want to understand how the computation is producing the
• 37:32
  new, this is an aesthetic inquiry. But then at the same time I was dealing with a sort of problem,
• 37:38
  a theoretical challenge that I found like contemporary debates in and on digital aesthetics to
• 37:45
  be stuck on. They were stuck on this sort of an impasse, I called it in previous work, between like
• 37:52
  the supposed continuity of, you know, lived experience and sensation, and perception, and the things
• 37:58
  that aisthesis is is really about, and then the discreteness the computational technologies are
• 38:05
  all about. Okay pixels, and instructions, and numbers, and functions, as discrete things ... but also the
• 38:13
  kind of discreteness we were talking about just moments ago. The discreteness inscribed in the logical level
• 38:19
  of an algorithmic procedure or what is a step-by-step procedure. So there was this big clash between
• 38:24
  continuity and discreteness. And the book came out in 2018, but I've wrote it, like in many years... it was a
• 38:33
  long project, it was in fact my doctorate, my doctorate project, and what I was observing you know
• 38:40
  when you're doing your doctorate you also start to start from a problem in the field, and
• 38:44
  the problem in the field, in media philosophy or philosophy of technology, was for me this clash, the
• 38:50
  fact that when people wanted to talk about computers aesthetically, in this philosophical sense,
• 38:55
  they were making computers be what computers are not, so they were trying, like turning the
• 39:02
  quantitative into the qualitative, or trying to transduce to use a notorious Simondonian term,
• 39:09
  something that is quite extensive, informational, instead in something that is intensive,
• 39:15
  forceful and affective. And for me that wasn't really speaking about the specificity of computation.
• 39:21
  So, a digital aisthesis, for me, it's really a contradiction in terms, it's like something that, in fact,
• 39:29
  is impossible. So I to find a way to reconciliate the digital, in the sense of the discreteness,
• 39:38
  with the aesthetic, in the sense of the continuity of a procedure. And that’s really
• 39:46
  the two reasons why I had to situate contingent computation, the concept of contingent computation, within
• 39:52
  aesthetics. At the same time, I must say that, in a sense, in a way sometimes I joke that I have said
• 39:59
  everything I had to say about aesthetics. Because
• 40:06
  it was for me an entry point to build an ontology of computation and because my ontology of computation
• 40:14
  had to be foundational, we had to start from these questions. But I don't see my own
• 40:20
  philosophical project as being limited to aesthetics. In a way, I've done aesthetics. Now I'm doing
• 40:26
  different things because I'm building different parts of the system. I’m a very systematic
• 40:30
  thinker, so, okay, done aesthetics. Now I have to do, like, epistemology ... but everything is
• 40:35
  really ontology at the end because I'm a metaphysician at the core. So, computation, unique you know
• 40:45
  modes and aspects of contingency and meshed in the production of ontological novelty, it's
• 40:50
  coupled to thought and experience, but has kind of its own domains but there's some serious
• 40:54
  enmeshment and some touching... so like super abstract question like what does computation do for the
• 41:02
  human and to the human? Like what are the kind of downstream implications for us as humans, kind of living
• 41:10
  cooperatively with computation and what are the specific affordances that computation you know
• 41:18
  provides us as humans? Super, super general question. On an intellectual and speculative level I think
• 41:24
  the greatest gift of studying computation is really like engaging with the question of
• 41:28
  what thought is, what thinking is, and I don't mean it like in the confined sense of cognition, of
• 41:36
  intelligence, but more broadly about really how thought is eventuated, how it is an event
• 41:46
  in the world and in reality, and what is the relationship between thought and world. So,
• 41:52
  obviously, as everybody working on on these topics have done a lot of work on artificial
• 42:00
  intelligence, like I've been doing that for for 20 years, but for me the question really that
• 42:06
  computation gives us in relation to that, and to also understand the human and and what the human
• 42:12
  is and does, is that it helps us to reformulate the question of thought in terms of not so much
• 42:19
  "who thinks?" but "what is thought?" Let me explain this better. In the sense that I don’t...
• 42:27
  when addressing artificial intelligence, for instance, I'm less interested in
• 42:32
  try to identify whether a machine thinks in the way in which a human can think, or a dog can think,
• 42:40
  or a plant can think, or a system can think... but I rather take the question more from like the ground
• 42:48
  So, “What is thought?” It's not, like, “Does this think like me?”, it's more, like,
• 42:56
  “Does this have thought?” And if it has got any form of thought, what kind of thought it is, and how can I even
• 43:03
  possibly recognize this as thought is that thought is so incommensurable with my own
• 43:08
  phenomenological ways of addressing reality. I'm saying phenomenological ways not because I'm
• 43:14
  phenomenologist, I'm really not, but because we are limited to our own phenomenology in order to
• 43:20
  address any sort of thing in the world. So this system that it's out there that we call it
• 43:27
  computation, these process, really for me, the greatest, like, intellectual gift is helping, like,
• 43:34
  to really rethink what thought is in a way starting from scratch. On a practical level,
• 43:43
  then, I'm also very much invested in addressing technology in non-prosthetic terms.
• 43:50
  So as not an augmentation, extension of the human, but something that
• 44:01
  exists alongside us, in proximity to us, with us and does things together with us, but not as an
• 44:10
  extension of the human. So, I'm definitely an advocate for a non-prosthetic understanding of
• 44:20
  technology. That's also why, to an extent, I find some projects such as post-humanism, and other ways
• 44:28
  to relate the human and the technological, ultimately disappointing, because in my opinion they always
• 44:35
  tend to default in a kind of prosthetic understanding of the technological — the cyborg being obviously
• 44:45
  the epitome of the prosthetic relation. And I'm generalizing here of course, for the sake of time,
• 44:51
  but that's generally what I think it's happening there. So, I think that the
• 44:58
  challenge is to try to think the alliance between the technological and the human. We have
• 45:05
  to forge an alliance of some sort. The ground upon which to build this alliance might be the
• 45:12
  question of thought, not because machines will take our jobs or take over the world, but because
• 45:20
  that's really the inherent ground upon which both projects, the technological the human, are built.
• 45:27
  At the same time, this alliance has to be forged by first recognizing the fracture between the human
• 45:37
  and the technological in the sense we are … yes, that there is a co-determination of the technological
• 45:43
  the human, yes, I think that's generally correct but at the conceptual level, at the
• 45:48
  level of the concepts and the abstractions that we need to build in order to engage with the world
• 45:53
  that will come. We need to start from the fracture, from the differences, and forge an alliance
• 46:00
  upon those recognized specificities, or at least that's been my project so far. Okay I might have
• 46:08
  a wild final question that might not make sense, but since we can edit things… If we think of
• 46:16
  computation or anything as a kind of from a historical perspective, I guess, the question would be
• 46:21
  what is the future of computation? The future of computation is the future of thought itself, in a way, but not
• 46:28
  because… I'm not here subscribing to any computational theory of mind, it's not because the mind is a computer,
• 46:35
  but because, as I was saying before, it's an eventuation of a form of thought that has to be
• 46:42
  thought alongside other forms of thought, and the future of computation will be the future of
• 46:48
  this concept as well, and and vice versa. Although thought is a broader category than the computational,
• 46:55
  so computation is a form of thought, but not all thought is computation. I’m definitely not making
• 47:01
  the claim, I'm not a computationalist in that sense. So the future of computation is also a future that,
• 47:09
  in a way, has to be conceptually constructed, and through those constructions, those conceptual
• 47:16
  constructions, we also need to engage with the limits and possibility of the computational.
• 47:23
  I've written a lot in the past about the concept of metacomputation, which is the concept, really...
• 47:30
  before we talked about my definition of computation as a method of systematizing
• 47:35
  and arranging reality through finite, logical quantitative means. And that's fine, and that's good, but
• 47:44
  the problem is, or the issue rather is, how we will understand these systematizations and these
• 47:49
  organizations and construction the computation builds. If we understand them as
• 47:54
  total matrix of all-enframing capture and reduction, then we are in big trouble.
• 48:01
  Because, like first of all, "there's nothing outside computation” — that's wrong, I think
• 48:07
  there are many things outside the computational. And at the same time, it's also politically,
• 48:12
  ethically, quite problematic. I mean that's metacomputation was really Leibniz’s project of
• 48:20
  reducing everything that can be said thought into like an organizational schema, like
• 48:28
  mathesis universalis, what he called it. So yes, it's Leibniz's project, but it's also Silicon Valley's project...
• 48:38
  there is a convergence there. And, interestingly, when philosophy criticizes computation (and
• 48:45
  philosophy has to criticize computation) it criticizes computation on the assumption,
• 48:52
  on the ground, or on the basis, that computation cannot be anything other than this total enframing,
• 48:59
  epistemic reduction of everything. Although the book “Contingent Computation" is not overtly
• 49:06
  political, there is a really kind of quite radical gesture in that, because it's shown a way
• 49:11
  in which, like a speculative way, in which we can understand computation otherwise... not as this
• 49:20
  total enframing of nature, life, experience, economy, and so on and so forth, but as something that
• 49:27
  can be indeed speculated about otherwise. So yeah, the future of computation — it's open but generally,
• 49:37
  I tend to believe the most futures are open. Thank you so much to M. Beatrice Faze for joining us!
• 49:47
  You know, a friendly reminde,r this is a 1 billion percent bootstrapped project without any ads or
• 49:53
  subscription revenue. It’s... it's a lot of work. Tell your friends about this podcast and perhaps more
• 50:00
  helpfully, you know, leave a little rating and review. We’re starting to enjoy a healthy degree of, you
• 50:06
  know, presence and we're really like for real emboldened by the support like it's it's actually
• 50:12
  really really meaningful to us so thanks, thanks everybody! Also, we're welcoming
• 50:19
  Helena McFadzean to the crew: a curator, write,r and heavy thinker with London roots based in Vienna.
• 50:25
  You'll definitely be hearing from her here soon. Thanks y’all.
• 50:36
  Subscribe to this podcast for more analysis at the intersection of algorithm subjectivity and the
• 50:47
  arts. Disintegrator Podcast.
• 50:55
  [Music]
• 50:57
  [Bell]