Temporary Ground - Abduction and Decision Making / EZE, 2026

On a Philosophy of Decision Making, A Response to Matt Grawitch

What falls short of the pragmatic in decision making as a philosophy?

Be Very, Very Quiet… I’m Hunting Fallacies

 The dynamics of decision making often calls for a temporary grounding 

in order to test out a conjecture, but at least a few implications are here: 

1)  the soundness of the conjecture as is  and  2) the implicit promise to evaluate the test and to adjust the premise ... .

The Illusion of Reason 

These problems also reveal themselves in the current discussion around AI decision making, the main discussion centering on the scope and consequence of a conjecture when implemented, but when consequences are virtualized . . .  , opportunities arise not only for the better evaluation of conjectures, but also for the fulfillment of testing and of subsequent adjustment.

History is usually different from reason, and this difference is often the site of the asemic.

AI on Pragmatic Decision Making

Pragmatic decision-making is a practical, results-oriented approach that prioritizes workable solutions over theoretical ideals or abstract principles. It focuses on "what works" in real-world scenarios, weighing evidence, constraints, and consequences to achieve specific goals efficiently. This method emphasizes adaptability and flexibility to handle  uncertainty. 

Key Components of Pragmatic Decision-Making:

• Results-Oriented: The value of a decision is judged by its practical consequences.
• Realism: Decisions are rooted in the current, actionable reality, rather than perfect, idealistic outcomes.
• Flexibility & Adaptability: Pragmatists adjust their strategies based on experience and changing circumstances.
• Efficiency: It prioritizes the most feasible, effective, and often quickest, course of action. 

Pragmatic vs. Idealistic Decision-Making:

• Idealist: Rejects a job offer because it is not their "dream job".
• Pragmatist: Accepts a lower-level job to gain experience, income, and leverage for future opportunities. 

How to Make Pragmatic Decisions:

1. Define Realistic Goals: Focus on what is attainable now.
2. Evaluate Evidence: Use available data to understand constraints and risks.
3. Prioritize Feasibility: Choose actions that can be implemented with current resources.
4. Experiment and Adapt: Test solutions using short feedback loops and adjust as needed. 

Pragmatic decision-making is valuable for managing risks and ensuring sustainability in dynamic situations. 

AI on Pragmatism

Pragmatism is a philosophical tradition that defines the meaning and truth of ideas by their practical consequences. Instead of looking for absolute, "mirror-like" descriptions of reality, pragmatists treat concepts and theories as tools or instruments for problem-solving and guiding action. 

Core Principles

• The Pragmatic Maxim: Formulated by Charles Sanders Peirce, it states that to understand a concept, you must consider what "practical effects" the object of that concept might have. Your idea of those effects is your whole idea of the object.
• Truth as "What Works": William James famously argued that truth is not a static property; it "happens" to an idea when it is verified through experience. A belief is true if it is "good in the way of belief"—meaning it helps us navigate the world successfully.
• Fallibilism: Pragmatists believe that no human knowledge is absolute or beyond the possibility of error. All beliefs should be treated as "working hypotheses" that may be refined as new evidence emerges.
• Inquiry over Spectatorship: John Dewey rejected the "spectator theory of knowledge," arguing that we are active participants in the world. Knowing is a process of inquiry—using intelligence to transform a confused situation into a resolved one. 

The "Big Three" Founders

Philosopher	Key Focus	Contribution
C.S. Peirce	Science & Logic	Founded the movement; emphasized using the scientific method to clarify meaning.
William James	Psychology & Morality	Popularized the movement; argued that beliefs (including religious ones) are justified if they provide practical value to a person's life.
John Dewey	Education & Democracy	Applied pragmatism to social reform; championed "learning by doing" in schools.

Pragmatism in Everyday Life

In a modern context, being "pragmatic" means choosing a sensible, realistic path over a purely idealistic or dogmatic one. 

• Example: Instead of spending hours arguing about the "perfect" way to split a restaurant bill (Idealism), a pragmatic group might just split it evenly to save time and enjoy the evening (Pragmatism). 

Would you like to dive deeper into how pragmatism influenced modern education or see how it compares to idealism?

AI on Abduction

Abduction, or "inference to the best explanation," is a generative reasoning process used to form hypotheses, where provisional (temporary) conclusions are adopted based on available evidence, subject to future revision. It works as a tentative foundation, where anomalies are observed, and hunches are tested to create a workable, yet non-final, understanding of a phenomenon. 

Key Concepts in Abduction and Temporary Foundations:

• Abduction Mechanism: It begins with observing a surprising fact, then inferring a hypothesis that, if true, would explain the fact. It is crucial for generating new ideas and theories, particularly in uncertain or complex situations.
• Provisional Foundations: The conclusions drawn through abduction are not considered absolute truths. Instead, they are "tentative," serving as a bridge to further investigation.
• Pragmatic Approach: Charles Sanders Peirce, who pioneered the study of abduction, argued that these foundations are grounded in practical action (pragmatism), adapted based on how they work in reality.
• Conditional Abduction: Abduction can be conditional, meaning a new, temporary rule is created to make sense of a situation, which is then formally tested.
• Structure of Reasoning: The process often involves four steps: observing anomalies, generating hunches, creating a deductive structure, and finally testing it through induction. 

Applications:

• AI and Modeling: Abduction serves as a foundation for artificial intelligence to create mental models and understand the world, particularly in creating agent-based systems.
• Research and Theory: It is essential for researchers to develop new theories when existing ones cannot explain new observations.
• "Inference to the Best Explanation": Known as IBE, this method selects the most plausible explanation from a set of candidates, which serves as a temporary, "best" foundation, for example as detailed in this Wikipedia page on Abduction. 

 AI Overview of Epistemology

Epistemology is the branch of philosophy focused on the nature, origin, scope, and justification of knowledge. It investigates what knowledge is, how we acquire it, and the difference between justified belief and opinion. Key questions include: What is knowledge? How do we know what we know? And, what can we know?. 

Etymology
Derived from the Greek words epistḗm(ē), meaning "knowledge," and -logy, which means "study of" or "discourse". 

Key Questions and Concepts

• Definition of Knowledge: Traditionally viewed as justified true belief (JTB), though this is debated, especially following Gettier problems.
Justification: The reasons or evidence supporting a belief.
Internalism vs. Externalism: Internalists argue that justification is based on internal mental states, while externalists believe external factors (like reliable processes) can justify knowledge.
The Scope of Knowledge: Investigating the limits of what can be known, including problems regarding reality and knowledge of other minds. 

Main Branches/Types of Epistemology

• Propositional Knowledge: Focuses on knowledge of facts ("knowing that...").
• Acquaintance Knowledge: Knowledge gained through direct experience or familiarity (e.g., knowing a person or place).
• Procedural Knowledge: "Know-how" or understanding how to perform tasks.
• Virtue Epistemology: Focuses on the intellectual virtues of the person (the agent) seeking knowledge.
• Epistemic Consequentialism: Suggests that ways of forming beliefs are justified if they lead to true beliefs.
• Common Research Approaches: Include Positivism (empirical observation), Interpretivism (human understanding), and Pragmatism (practical application). 

Famous Epistemologists

• Plato: Explored knowledge as justified true belief.
• René Descartes: Focused on doubt and foundational knowledge.
• John Locke & David Hume: British Empiricists focusing on experience.
• Immanuel Kant: Bridged rationalism and empiricism, developing the analytic-synthetic distinction.
• Edmund Gettier: Challenged the traditional JTB definition.
• Bertrand Russell: Distinguished between knowledge by acquaintance and description.

flight / EZE, 2026

 

Asemic Displacement - AI / EZE, 2026

Asemics Is about ( )-Placed Meaning

Asemic writing is a sub-category of asemics, but AI and AI writing are  also sub-categories of asemics.

Asemic writing promotes an understanding of writing as, largely, something other than writing.

AI promotes an understanding of intelligence as something other than intelligence based on understanding See  Intelligence without Understanding.

Asemic writing is about the displacement of writing onto art. It often takes the patterns within writing and presents them as art.

AI is a tool that displaces understanding; AI is also a tool to locate meaning within certain groupings without recourse to understanding the particulars, the groupings, the relationships, or anything at all. Indeed, AI assimilates patterns that allow it to simulate an understanding of those patterns.

Asemic writing does not need to preserve understanding within the bounds of writing as its meaning is often beyond writing, in art.

AI writing promotes understanding without recourse to understanding, but to high-percentage patterns.

Asemic writing and AI writing are much the same in terms of patterning and in the use of patterns.

But what is also common to asemic writing, AI, and AI writing is the ()-placement of understanding, i.e., asemic writing, AI, and AI writing often displace meaning onto a non-immediate context. This non-immediate context  is the displacement that asemics puts into play.

Asemics in Play

Asemic Understanding

Voynich Manuscript

AI Understanding

On AI Writing

AI Poetry

On Linkage

Contrastive Learning

AI:

Artists are increasingly using generative AI to intentionally produce asemic art by leaning into the machine's tendency to "hallucinate" text-like forms. Because AI does not inherently understand the semantics of language, it naturally produces symbols that mimic the shape of writing without the meaning, a state some call "inherently asemic" [1].

Specific AI Tools for Asemic Art

• Stable Diffusion & LoRA Models: Stable Diffusion is a favorite for asemic art because of its high customizability and open-source nature [1, 2]. Artists use LoRA (Low-Rank Adaptation) models to fine-tune the AI on specific datasets, such as 18th-century manuscripts or abstract calligraphy, to produce unique "textish" scripts [1].
• Midjourney: Known for its artistic stylization by default, Midjourney excels at creating evocative, moody compositions [1, 2]. While it often prioritizes aesthetics over strict text accuracy, artists use its "vibe" to generate surreal landscapes integrated with abstract symbols [1].
• GANs (Generative Adversarial Networks): Tools like StyleGAN allow artists to train the AI on their own previous artworks [1, 2]. By feeding the algorithm thousands of images of hand-drawn marks, the artist "teaches" it a specific aesthetic, and the AI then generates new, original variations [1, 2].
• Specialized "Gibberish" Tools: Simple programs like the Gibberish Generator allow users to mimic handwriting styles using basic settings to create "wordless" scripts [1].

Techniques to "Provoke" Asemic Hallucinations

1. Iterative Prompting: Artists often use prompts that emphasize abstract textures or "unreadable ancient manuscripts" to force the AI into generating symbols rather than legible words [1, 2].
2. In-painting and Latent Editing: In Stable Diffusion, artists can use inpainting to "wipe" a specific area and tell the AI to fill it with "abstract calligraphic noise," effectively curating where the hallucinations occur [1, 2].
3. Cross-Genre Fusion: By prompting for a mix of "Western oil painting" and "Eastern ink wash," artists can trigger the AI to blend visual languages into entirely new, non-signifying hybrid styles [1].
4. Low Guidance (CFG Scale): Lowering the Classifier-Free Guidance (CFG) scale in tools like Stable Diffusion allows the AI more "creative freedom," often leading to more surreal, hallucinatory outputs that depart from literal instructions [1, 2].

dotly / EZE, 2026