## Metadata - Author: [[Thomas Bartscherer and Roderick Coover]] - Full Title: Switching Codes - Category: #books ## Highlights - It is perhaps inevitable that a book focusing on the impact of digital technology will, at some deep level, raise questions about the status of human being in the world. In a traditional conception of human being stretching back to Aristotle, the human is located between the bestial and the divine. In modernity that traditional notion has been challenged from both sides; the line separating human from animal has become ever more difficult to draw, while at the same time the divine has for many ceased to be a relevant category for philosophical thought. While these two categories, conceived as sub- and supra-human respectively, have certainly not disappeared from the discourse, the present age demands that we add a fourth term to the Aristotelian triad. Now we must also consider the being of humans relative to the being of machines. ([Location 139](https://readwise.io/to_kindle?action=open&asin=B00AKSY148&location=139)) - we believe that the traditional codex nurtures modes of thought and being that are increasingly rare in our time. Festina lente—the proverb that the great Renaissance publisher Aldus Manutius chose for his imprint—is now, more than ever, the most fitting motto for book culture. To make haste slowly means to cultivate the habits of mind and hand that attend the writing, publishing, and reading of books. ([Location 158](https://readwise.io/to_kindle?action=open&asin=B00AKSY148&location=158)) - sounds like tbl [Tim Berners-Lee] circa 1999. massive linking is exactly what scientists have tried to do for centuries. “Tables of contents, indexes, bibliographies, and reference sections are hypertext links. . . . Suddenly scientists could escape from the sequential organization of each paper and bibliography, to pick and choose a path of references that served their own interest. ([Location 1475](https://readwise.io/to_kindle?action=open&asin=B00AKSY148&location=1475)) - Scholarship, you could say, is the capacity to analyze the same object with different criteria, and different objects with the same criteria. ([Location 1485](https://readwise.io/to_kindle?action=open&asin=B00AKSY148&location=1485)) - Not too long before the field of artificial intelligence was born, there were some who felt the field of mathematics was splintering into too many subareas and that it had lost touch with the real world. One of those whose work has been seen as an important contribution to unifying the field, and to bringing the field of applied mathematics back to the fore, is John von Neumann, one of the fathers of modern computing. Reflecting on this accomplishment, he observed: Mathematical ideas originate in empirics. . . . But, once they are so conceived, the subject begins to live a peculiar life of its own and is better compared to a creative one, governed almost entirely by aesthetical motivations. . . . As a mathematical discipline travels, or after much abstract inbreeding, [it] is in danger of degeneration. . . . Whenever this stage is reached, the only remedy seems to me to be the rejuvenating return to the source; the reinjection of more or less directly empirical ideas. (1947, 195) I strongly believe that a critical step for the future of knowledge representation is for us to apply von Neumann’s approach to mathematics to our field. ([Location 2450](https://readwise.io/to_kindle?action=open&asin=B00AKSY148&location=2450)) - No one knows the number of concepts involved in human thought, and early estimates of thousands and later millions have now given way to numbers in the billions—but we are all just guessing. On the Semantic Web we’re also guessing, but starting from a much higher number—some startups are talking about numbers in the billions, and that’s just representing the data instances. When we look at ontologies per se, that is, documents where there is at least some definition of classes (as opposed to just instance data), Google finds over eighteen thousand files with the extension “owl,” the great bulk of which appear to be ontology files in the web language OWL. Semantic search engines currently report over two million web files that contain information in RDF, RDF Schema, and OWL. ([Location 2482](https://readwise.io/to_kindle?action=open&asin=B00AKSY148&location=2482)) - the information in these ontologies and knowledge bases is often contradictory and even occasionally formally inconsistent. As an example, a crawl of the Semantic Web retrieves several documents that purport to include my age. Since these have originated over a decade, you can find my age to be anything from thirty-nine to “approaching fifty,” to “I refuse to answer that question.” Obviously these cannot all be true, and any reasoner that is unable to deal with inconsistent knowledge is going to be doomed if it tries to play on the web. ([Location 2501](https://readwise.io/to_kindle?action=open&asin=B00AKSY148&location=2501)) - most traditional KR systems are based on first-order logics where (P AND NOT P) IMPLIES Q. Again, welcome to the real world. When we have disagreement, error, and deceit (all of which are hallmarks of systems with large numbers of humans pursuing multiple goals), we’d better have a more robust infrastructure. ([Location 2518](https://readwise.io/to_kindle?action=open&asin=B00AKSY148&location=2518)) - artistic creation has to be simultaneously unpredictable, harmonious, and familiar. ([Location 2587](https://readwise.io/to_kindle?action=open&asin=B00AKSY148&location=2587)) - It is obvious that the logical inferences involved in creative abilities cannot be reduced to mere deduction (i.e., to inferences from the general to the particular), because deduction is by nature conservative, whereas creation corresponds to an increase in knowledge. The word “create,” after all, comes from the Latin verb crescere, which means to grow; in other words, the output of a creative process has to contain more than was given as input. ([Location 2597](https://readwise.io/to_kindle?action=open&asin=B00AKSY148&location=2597)) - One of the most insidious traps of the cognitivist/logicist camp is the representational flatland of verbal models. All knowledge, reasoning, representation, decision making, and so on are viewed as transformations of linguistic expressions: speech, conceptual maps, semantic networks, predicate calculus, and so on. Even interpreting diagrams is reduced to verbal manipulations (Clancey 2000). ([Location 3158](https://readwise.io/to_kindle?action=open&asin=B00AKSY148&location=3158)) - A badly reproduced original, we will argue, risks disappearing, while a well-copied original may enhance its originality and continue to trigger new copies. Facsimiles, especially those relying on complex (digital) techniques, are thus the most fruitful way to explore the original and even to redefine what originality is. ([Location 4967](https://readwise.io/to_kindle?action=open&asin=B00AKSY148&location=4967)) - let us remember that the word “copy” need not be derogatory; indeed, it comes from the same root as “copious,” and thus designates a source of abundance. A copy, then, is simply a proof of fecundity. ([Location 4970](https://readwise.io/to_kindle?action=open&asin=B00AKSY148&location=4970)) - to be original means necessarily to be the origin of a lineage. That which has no progeny, no heirs, is called, not original, but sterile, barren. ([Location 4974](https://readwise.io/to_kindle?action=open&asin=B00AKSY148&location=4974)) - No matter how mechanical a reproduction is, once there is no huge gap in the process of production between version n and version n + 1, the clear-cut distinction between the original and its reproduction becomes less crucial—and the aura begins to hesitate and is uncertain where it should land. ([Location 5070](https://readwise.io/to_kindle?action=open&asin=B00AKSY148&location=5070))