Temporal Ethics and Chronomorphic Cognition: Identity, Truth, and Frame Memory in Artificial Intelligence
Fuente:
Zenodo
Saved in:
| Main Author: | |
|---|---|
| Format: | Recurso digital |
| Language: | English |
| Published: |
Zenodo
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866901610000547840 |
|---|---|
| author | Kevin Fathi |
| author_facet | Kevin Fathi |
| contents | <p>Hallucination in generative AI systems is often treated as a technical flaw—an<br>artifact of insufficient training data or suboptimal fine-tuning. In this paper, we argue<br>that hallucination stems from a deeper cause: the absence of temporal perception.<br>Drawing on Gregory Bateson’s theory of recursive framing, we extend the Batesonian<br>AI framework to model time not as linear metadata but as a sequence of evolving<br>interpretive frames. We define temporal frames as recursive structures that track an<br>agent’s shifting stance, memory, and contradictions across time. Using this model, we<br>propose a novel agent architecture capable of detecting frame drift, resolving temporal double binds, and preserving narrative coherence. Simulation scenarios are presented in which agents retell stories, self-correct false beliefs, and manage temporally layered instructions. We introduce the concept of chronomorphic cognition—an AI’s ability to think through time—as a requirement for epistemic integrity and long-term coherence. Our findings suggest that solving hallucination will require not just better data, but a rethinking of time itself as a recursive interpretive process.</p> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_15179908 |
| institution | Zenodo |
| language | eng |
| publishDate | 2025 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | Temporal Ethics and Chronomorphic Cognition: Identity, Truth, and Frame Memory in Artificial Intelligence Kevin Fathi AI hallucination recursive learning frame theory temporal consistency Batesonian AI chronomorphic cognition artificial intelligence ethics machine memory narrative coherence multi-agent reasoning <p>Hallucination in generative AI systems is often treated as a technical flaw—an<br>artifact of insufficient training data or suboptimal fine-tuning. In this paper, we argue<br>that hallucination stems from a deeper cause: the absence of temporal perception.<br>Drawing on Gregory Bateson’s theory of recursive framing, we extend the Batesonian<br>AI framework to model time not as linear metadata but as a sequence of evolving<br>interpretive frames. We define temporal frames as recursive structures that track an<br>agent’s shifting stance, memory, and contradictions across time. Using this model, we<br>propose a novel agent architecture capable of detecting frame drift, resolving temporal double binds, and preserving narrative coherence. Simulation scenarios are presented in which agents retell stories, self-correct false beliefs, and manage temporally layered instructions. We introduce the concept of chronomorphic cognition—an AI’s ability to think through time—as a requirement for epistemic integrity and long-term coherence. Our findings suggest that solving hallucination will require not just better data, but a rethinking of time itself as a recursive interpretive process.</p> |
| title | Temporal Ethics and Chronomorphic Cognition: Identity, Truth, and Frame Memory in Artificial Intelligence |
| topic | AI hallucination recursive learning frame theory temporal consistency Batesonian AI chronomorphic cognition artificial intelligence ethics machine memory narrative coherence multi-agent reasoning |
| url | https://doi.org/10.5281/zenodo.15179908 |