Particularized Truth and Reality Testing in the Algorithmic Era
How can two individuals posing the same question to a conversational model receive different answers, and what are the implications for our understanding of truth and reality testing? The human search for truth has taken various forms: truth as revealed, investigated through science, encountered psychologically, and, today, approached through probability and computation. We distinguish between truth, relative truth, and reality testing, proposing that reality testing relies on an “outside” grounded in objective criteria. Yet conversational models do not merely retrieve information; they formulate a representation of reality within an encounter with a specific inquirer. The user’s input and history participate in shaping the response, such that the “outside” returns tailored to their world. The novelty lies not in algorithmic mediation as such, but in the structural shift in the relationship between subject, representation, and outside.We propose the concept of Particularized Truth: neither a new ontological category nor a mere adjustment of style, but a form in which a claim about reality becomes present for a specific subject within an algorithmic encounter, influencing content and recommendations. Research on sycophancy provides empirical support for aspects of this adaptive process. Under some conditions, such adaptation may shift the interaction from the demands of the reality principle toward the pleasure principle—the wish that the world offer no resistance. We therefore propose an expansion of reality testing in the AI era, examining how representations are generated and their capacity to diverge from the user’s existing perspective. Reality testing requires an outside, even when the outside has learned to speak our language.
Authors
- Dror Oren (ORCID: https://orcid.org/0000-0003-3117-4065)
- Daniela Ben Zeev
Publication Details
- Journal
- PsyArXiv (OSF Preprints)
- Published
- 2026-11-15
- DOI
- https://doi.org/10.31234/osf.io/m423e_v1
- Primary Topic
- Cybernetics and Technology in Society
- Type
- preprint