Research

spatial intelligence

I’m researching spatial intelligence. Both humans and machine learning models interpret 3D space differently. Human spatial expectations recalibrate to the spatial patterns we repeatedly encounter — including synthetic ones. In this context, custom projects can utilize image and video packages to enhance understanding. AI will change us, and we will change it, which isn’t necessarily bad; it’s just different. Exposure leads to familiarity, which in turn drives an expectation shift in our visual execution.

working title

Inferential Normalization & Hybrid Continuity:
How Modalities & Constraints Shape Spatial Representation

how spatial structure gets decided across modalities, constraints, and human steering

Abstract cosmic swirl with vibrant colors and glowing lines.

Causal Framework

1 → working theory: inference becomes reality

externalization → adaptation → redesign

2 → data shapes experience

4 → 3D data to world hypothesis

2 → data shapes experience

In generative media, that world-model is built from explicit modalities and constraints; different modality bundles and coordination produce different coherence signatures. For example, in custom projects involving image and video packages, outputs can appear coherent while failing to preserve continuity, identity, scale, or geometric stability in the way ecological perception would. The key variable is not the tool; it’s whether structure is being specified or resolved at that step, especially during visual execution.

3 → how data interacts

4 → 3D data to world hypothesis

2 → data shapes experience

My core research question is how each modality supplies different spatial cues and constraint conditions, how those constraint sets interact, and how model architecture (alignment vs separation) determines what spatial structure can stay consistent—especially when human steering functions as an added modality during generation. This inquiry is particularly relevant in the context of a custom project focused on developing image and video packages, where visual execution plays a crucial role in maintaining that spatial consistency.

4 → 3D data to world hypothesis

Method: input modality → constraints → constraint interaction → alignment/separation → architecture-shaped structure → representation in custom projects → human interpretation of image and video packages through visual execution.

5 → why this matters

6 → dissertation outcome

4 → 3D data to world hypothesis

Humans rely on spatial structure to reduce interpretive effort and support action, particularly when engaging with custom projects that involve complex image and video packages. Teaching people how to reason about inferred worlds reframes imperfections as intelligible outcomes rather than failures, thus enabling effective communication, adaptation, and intentional intervention in hybrid human–AI systems, which can greatly benefit from thoughtful visual execution.

6 → dissertation outcome

Deliverable: a repeatable casework framework for inferential normalization and world-model shaping that links a system’s modality bundle and coordination to the coherent-enough spatial hypothesis it constructs—and to the interpretations humans reliably take from it in hybrid environments. This custom project includes a documentation workflow (inputs → human interventions → breakpoints → constraint dynamics mapped) that makes hybrid continuity observable, comparable, and archivable over time, while also integrating image and video packages for enhanced visual execution.

6 → dissertation outcome

track A: specified-heavy workflows (explicit 3D authoring + ai modifiers) that can be utilized in a custom project.

track B: resolved-heavy workflows (generative worldbuilding) that support the development of image and video packages for effective visual execution.

NEXT STEPS

**operationalize framework in real tool sessions**
run recorded workflows, capture decision points, and map each event while refining the documentation protocol.

**characterize each platform as a coordination profile**
produce tool profiles that specify available modalities, dominant constraint biases, intervention levers, and recurrent coherence/failure signatures (**lens: specified ↔ resolved**).

**execute comparative modality experiments across tools**
run controlled probe families that vary modality bundles, prompt completeness, and coordination conditions to identify cross-tool regularities and divergences in inferred spatial structure.

A monochrome sculpture of a performer on stage with a crowd and spotlight. — **synthesize into an empirical account of inferential normalization**
connect observed coherence signatures to human interpretation and expectation shift, then formalize the archive and chapters that demonstrate hybrid continuity as evidence-backed.

annie burke

Research

spatial intelligence

working title

Causal Framework

1 → working theory: inference becomes reality

externalization → adaptation → redesign

2 → data shapes experience

4 → 3D data to world hypothesis

2 → data shapes experience

3 → how data interacts

4 → 3D data to world hypothesis

2 → data shapes experience

4 → 3D data to world hypothesis

4 → 3D data to world hypothesis

4 → 3D data to world hypothesis

5 → why this matters

6 → dissertation outcome

4 → 3D data to world hypothesis

6 → dissertation outcome

6 → dissertation outcome

6 → dissertation outcome

6 → dissertation outcome

6 → dissertation outcome

NEXT STEPS

Annie Burke: Digital Artist

Research

spatial intelligence

working title

Causal Framework

1 → working theory: inference becomes reality

externalization → adaptation → redesign

2 → data shapes experience

4 → 3D data to world hypothesis

2 → data shapes experience

3 → how data interacts

4 → 3D data to world hypothesis

2 → data shapes experience

4 → 3D data to world hypothesis

4 → 3D data to world hypothesis

4 → 3D data to world hypothesis

5 → why this matters

6 → dissertation outcome

4 → 3D data to world hypothesis

6 → dissertation outcome

6 → dissertation outcome

6 → dissertation outcome

6 → dissertation outcome

6 → dissertation outcome

NEXT STEPS

Annie Burke: Digital Artist

This website uses cookies.