ChatGPT for Technomancy

The first time I ever presented publicly about cybernetics and sorcery was at the Ekstasis! seminars (“Magical Seminars for the Postmodern Mind”) hosted by the Illuminates of Thanateros in 2007. It was during my research for that presentation, “Re/Considering the Cybernetic Model of Magic,” that I first encountered conceptual blending theory via Jesper Sørensen’s A Cognitive Theory of Magic. Since then, the basic ideas of conceptual blending have contributed significantly to my technomantic innovations, from the “Beginning Interactive Multimedia Ritual Design” course I taught online in 2010, which evolved into the Technomancy 101 tutorial I published in 2018, to the robomantic work I have been developing since before 2011. I even included a cognitive blending exercise in the Technomancy 101 curriculum. Regardless of conceptual blending’s merit as an expository model of cognition, I have found it endlessly rewarding as a generative utility for blending art, magic, and technology to produce novel technomantic designs. In the spirit of that creative potential, I want to say a bit about ChatGPT.

A few folks have been applying recent developments in artificial intelligence to their magical practice; see e.g. the works of Alley Wurds. Most of what I have witnessed in this space is people generating images from DALL-E, Midjourney, and the like for use in enchantment, divination, &c. I have also seen ChatGPT called on to suggest ideas for ritual activities. All valid examples of technomancy (I shall not write “legitimate” because technomancy remains a totally bastard art, thank gods), however, here I want to focus on ChatGPT’s ability to apply conceptual blending to the generation of technomantic designs. One could jump straight to soliciting new designs from ChatGPT, but I dig a robust theoretical framework.

First, let’s see what ChatGPT knows about conceptual blending (I am including its full reply here for readers who are unfamiliar with CB):

Yes, I'm familiar with conceptual blending theory. Conceptual blending, also known as conceptual integration, is a theory in cognitive science that explains how individuals combine various elements from different concepts to create new ideas or understandings. This theory was developed by Gilles Fauconnier and Mark Turner and has been influential in fields such as linguistics, psychology, philosophy, and artificial intelligence.

Key points of conceptual blending theory include:

    Mental Spaces: The theory suggests that our minds form 'mental spaces' — partial structures that represent situations or elements from the real world or imagination. These spaces are dynamic and context-dependent.

    Cross-Space Mapping: Elements and relationships from different mental spaces can be projected onto a new space, known as the 'blend.' This blend integrates elements from the original spaces to generate new meanings, ideas, or interpretations.

    Selective Projection: Not all elements from the input spaces are incorporated into the blend. Instead, the process is selective, often based on relevance, salience, or the individual's intentions and background knowledge.

    Emergent Structure: The blend can develop emergent properties, meaning it can contain structure and meaning not explicitly found in any of the input spaces. This emergent structure allows for creative thinking, problem-solving, metaphorical reasoning, and the generation of new ideas.

    Iteration and Recursive Blending: Blending is not a one-time process but can be iterative. A blend can serve as an input space for further blending, allowing for complex chains of thought and reasoning.

Conceptual blending theory is particularly important in understanding how humans engage in creative thinking, how we understand and use metaphors, and how we can combine seemingly unrelated concepts to produce innovative ideas. It's also relevant in the study of language and discourse, as it provides insight into how meaning is constructed and communicated. In artificial intelligence, the principles of conceptual blending are explored to enhance the capability of AI in creative problem-solving and understanding human communication.

Not bad! Making a mental note to later see how recursive I can get with ChatGPT regarding that last sentence, I request an example of cross-space mapping:

Cross-space mapping in conceptual blending theory involves creating connections between elements from different mental spaces to form a new, integrated conceptual space. Here's an illustrative example:

Input Spaces:

    Space A - The Solar System: Consider a mental space where we think about the solar system. This space includes the sun, planets, moons, asteroids, and other celestial bodies orbiting the sun.
    Space B - A Busy City: Another mental space is that of a busy city. This space includes elements like roads, vehicles, traffic lights, and pedestrians.

Cross-Space Mapping:

    Blend - A Model of the Solar System as a Busy City: In the blended space, we create a metaphorical model of the solar system using elements from a busy city. For example:
        The sun is represented as a large, central roundabout or hub.
        Planets are depicted as vehicles on different lanes around the roundabout, each following its own path (orbit).
        Moons become smaller vehicles or bikes accompanying the larger vehicles (planets).
        Asteroids are akin to pedestrians or smaller moving elements in the city.

In this blended conceptual space, we use familiar aspects of city traffic to understand and represent the structure and dynamics of the solar system. The orbits of planets become lanes in a city, and celestial bodies are transformed into elements of urban traffic. This metaphorical mapping allows us to grasp the complex astronomical concept using a more familiar and tangible context.

This example of cross-space mapping demonstrates how conceptual blending allows for creative interpretation and understanding by merging elements from disparate domains to form a coherent, novel concept.

OK, computer; I blame Musk’s Roadster for that one, but I can tell Chatty’s got the gist. Let’s forge ahead into the good stuff: Considering the idea of “technomancy” in the context of conceptual blending theory, what are some ways that cross-space mapping might explain the emergence of the concept of technomancy?

The concept of "technomancy," a blend of technology and magic, can be understood through conceptual blending theory by examining how elements from the distinct mental spaces of technology and magic are mapped onto a new, integrated conceptual space. Here's an exploration of how cross-space mapping might explain the emergence of the concept of technomancy:

Input Spaces:

    Space A - Technology: This mental space includes elements like computers, software, digital networks, robots, and advanced scientific principles. Technology here is characterized by logical systems, electronic devices, and real-world applications based on scientific understanding.

    Space B - Magic: The mental space of magic encompasses elements such as spells, enchantments, mystical artifacts, supernatural beings, and arcane knowledge. Magic is often seen as mysterious, bound to ancient lore, and operating outside the conventional laws of physics.

Cross-Space Mapping:

    Blend - Technomancy: In the blended conceptual space of technomancy, elements from technology and magic are combined to create a new, imaginative concept. Examples include:
        Enchanted Devices: Gadgets and devices that operate not only through electrical components but also through mystical enchantments. For instance, a smartphone that can cast spells or a computer that processes data using magical algorithms.
        Magical Interfaces: User interfaces or software that respond to magical gestures or incantations, combining the practicality of technology with the mystique of magic.
        Cyber-Wizards: Individuals who harness both advanced technological skills and magical abilities, symbolizing a fusion of a hacker and a sorcerer.
        Mythical Cybernetics: Implanting mystical artifacts as cybernetic enhancements, allowing for abilities that are both magical and technological in nature.

In this blended space, the logical, rule-based systems of technology are infused with the mysterious and often rule-defying aspects of magic. This fusion creates a rich, imaginative concept where technological advancements are perceived and interpreted through the lens of magical practices and vice versa.

The emergence of technomancy as a concept reflects a creative integration of two seemingly opposed domains, allowing for novel ideas and narratives. It shows how conceptual blending facilitates the creation of new meanings and understandings by combining elements from different realms of thought. In literature, gaming, and other forms of media, technomancy serves as an example of how blending theory can be applied to understand the evolution of complex and imaginative concepts.

That is a pretty darned good answer! Let’s chase one of those rabbits a little further down the hole:1 Please give three examples of how cross-space mapping could be applied to “a smart phone that can cast spells.” For each example, use different elements from smartphones and different elements from spell-casting.

The concept of a smartphone that can cast spells is a fascinating blend of technology and magic. Using cross-space mapping from conceptual blending theory, we can create various imaginative scenarios by integrating different elements of smartphones and spell-casting. Here are three examples:

    Gesture-Based Spell Activation:
        Smartphone Element: Touchscreen Gestures
        Spell-Casting Element: Hand movements used in spell-casting
        Blended Concept: In this scenario, the smartphone recognizes specific hand gestures made on its touchscreen as commands to activate certain spells. For instance, drawing a spiral on the screen could summon a wind gust, while tracing a heart shape might cast a love spell. This blends the intuitive touch interface of smartphones with the traditional idea of gestures in magic.

    Voice-Activated Magical Commands:
        Smartphone Element: Voice Command Technology (like Siri or Google Assistant)
        Spell-Casting Element: Verbal incantations
        Blended Concept: In this example, the smartphone operates through voice-activated commands, similar to modern virtual assistants, but instead of mundane tasks, it executes magical spells. Speaking specific magical phrases or incantations into the phone triggers various spells. For example, saying "Illuminate!" could make the phone emit a bright light, mimicking a lumos spell.

    Augmented Reality (AR) Enchantment:
        Smartphone Element: Augmented Reality Technology
        Spell-Casting Element: Visualization and manifestation of magical effects
        Blended Concept: Here, the smartphone uses its AR capabilities to create visible, interactive magical effects in the real world. By activating an AR app and pointing the camera at a space, users could see and interact with magical phenomena. For example, pointing the phone at a door could display an AR image of the door being sealed with a magical barrier, which, in the blended concept, would actually lock or unlock the door as if by a spell.

Each of these examples demonstrates how elements from the domains of smartphones and magical spell-casting can be creatively combined to form a novel concept. They reflect the power of conceptual blending in generating innovative ideas by merging different realms of knowledge and imagination.

How about asking ChatGPT to assist in manifesting one of those examples? Would you please write some code for an Android phone that will turn on the phone’s flashlight when the operator says “Illuminate!”?

ChatGPT kindly obliges:

Creating a simple Android app that turns on the phone's flashlight when the user says "Illuminate!" involves combining voice recognition with flashlight control. Below is an example of how you might start coding such an app in Java, a common language for Android development. Please note that this is a basic example and assumes you have some familiarity with Android app development.

Before you start, ensure that you have the Android SDK and Android Studio installed, and that you have set up a new Android project. This code also requires the necessary permissions for camera access and audio recording (for voice recognition), which need to be declared in your AndroidManifest.xml file:

[code omitted for length]

Here is a basic example in Java:

[code omitted for length]

This code sets up a basic Android activity with voice recognition capabilities. When the user says "Illuminate!", the app will attempt to turn on the device's flashlight. Keep in mind that this is a simplistic example. In a full application, you would need to handle more edge cases and potentially refine the voice recognition for accuracy.

Also, remember that using the device's hardware features like the camera and microphone requires handling user permissions appropriately, both in the manifest and at runtime, as per Android's requirements.

(I added the code to Pastebin, if you want to read it there.) Now, casting Lumos with my phone is not really the sort of thing I’m interested in, but it and the other two suggestions are not far from what I got up to 10 or 15 years ago when I started this gig, and speaking a magic word or phrase to elicit programmable lights to do something interesting while situated in a ritual context is something I might still do today. With smarter prompting, ChatGPT’s replies could converge on greater relevance to what I am doing more recently, and it makes inferences remarkably similar to ones I myself have made over the years.

Could you please relate the conceptual blending concepts of simplex, mirror, single scope, and double scope blends, to what you said earlier about cross-space mapping and technomancy?

Certainly! The concepts of simplex, mirror, single scope, and double scope blends are types of blends in conceptual blending theory. These can be related to the earlier discussion about cross-space mapping and the concept of technomancy. Let's explore how each type of blend can be applied to the idea of technomancy:

    Simplex Blends:
        Description: In a simplex blend, one input space provides a frame or structure, and the other offers specific content or details.
        Application to Technomancy: Consider a scenario where the structure of technological processes (like programming or electrical engineering) is the frame, and the details are magical spells or enchantments. For instance, a computer program that operates like a traditional spell, where code lines function as magical incantations. The technological structure frames the magical content.

    Mirror Blends:
        Description: Mirror blends involve two input spaces that are very similar and reflect each other. The blend emphasizes the similarities and downplays the differences.
        Application to Technomancy: Imagine comparing a magical ritual to a technological process, like a software update. Both processes involve a series of steps leading to a transformation or an outcome. The blend highlights the ritualistic aspects common to both, such as following specific sequences (in spells or algorithms) to achieve a desired result.

    Single Scope Blends:
        Description: In single scope blends, one of the input spaces (the "scope") provides the organizing frame and structure, while elements from the other space are integrated into this structure.
        Application to Technomancy: Take the concept of a smartphone (from the technology space) as the organizing frame. Integrating elements from spell-casting, such as voice commands that are actually magical incantations, creates a technomancy device where the phone's functions are controlled through magical phrases.

    Double Scope Blends:
        Description: Double scope blends involve two input spaces that contribute equally to the structure of the blend, creating a space that is markedly different from either input.
        Application to Technomancy: Here, elements of technology and magic are equally influential, creating a novel concept. For example, a world where technology and magic are indistinguishable, and devices are powered by a fusion of electrical energy and mystical forces. This blend results in a unique setting where tech-mages program spells and engineers design artifacts with both circuits and runes.

Each of these blending types offers a different way to understand the fusion of technology and magic in the concept of technomancy. They demonstrate how varying degrees of influence and integration from the input spaces can create diverse and imaginative outcomes in conceptual blending.

Again, not always the best examples at first, and the usual caveats regarding ChatGPT apply, of course;2 but you can see how it is quite capable of helping to navigate a blended space of generative ideas about technomancy, and we merely scratched the surface.3

Here are a few more books about conceptual blending if you are interested:

  • The Way We Think: Conceptual Blending and the Mind’s Hidden Complexities by Gilles Fauconnier and Mark Turner (Basic Books 2002)
  • Designing with Blends: Conceptual Foundations of Human-Computer Interaction and Software Engineering by Manuel Imaz and David Benyon (MIT Press 2006) — published the same year as Sørensen’s book, these two together presaged Hyperritual and Eldritech
  • Spaces of Interaction, Places for Experience by David Benyon (Springer 2014)

  1. This is an example of a conceptual blend where the concept of “following” a statement through further elaboration is combined with the concept of hunting a lagomorph, and has associations with Alice’s Adventures in Wonderland and The Matrix. ↩︎
  2. I have found ChatGPT increasingly reluctant to answer queries about real magic, and it is often factually wrong about any number of things, which does not trouble me because people are often factually wrong about any number of things. E.g., I was recently conversing with ChatGPT about H.P. Lovecraft’s “The Dreams in the Witch House,” and decided to test its comprehension by asking who the name Nahab refers to. It responded that there is no character named Nahab in the story, and suggested two other characters instead. I replied with a quotation from the story in which the name appears, after which ChatGPT amended its answer and even, unprompted, provided an explanation for the name’s significance. By the way, this is an example of what Heinz von Foerster calls an illegitimate question because we already know the answer to it. I am most interested in ChatGPT’s potential for answering legitimate questions i.e. ones no one yet knows the answers to. ↩︎
  3. “Scratching the surface” is another blend. ↩︎