What the study found
The paper argues that under finite capacity, the future is not empty possibility but an operational horizon already shaped by active traces, expectations, constraints, admissibility conditions, and witness availability. It introduces three related ideas: anticipatory trace, trace foreclosure, and horizon governance.
Why the authors say this matters
The authors suggest this matters because, in AI-facing and human-scale settings, a process can become less accountable if it closes, narrows, or replaces the trace-formation needed for reviewable continuation. They also state that horizon governance is a discipline for protecting the admissible space of continuation under finite capacity.
What the researchers tested
This paper is an orientation and bridge note within the Trace–Continuation under Finite Capacity (TCFC) series of the Synkyria Project. It builds on TCFC-01 and TCFC-02 and connects the proposed grammar to AI generation, learning, TER runtime evidence, SFV translation, and the “No AI before trace” principle.
What worked and what didn't
The paper presents the concepts of anticipatory trace, trace foreclosure, and horizon governance as additions to the TCFC framework. It also says that anticipatory trace is a future-oriented pressure that becomes active before the expected event occurs, and that trace foreclosure happens when witnessable trace is prematurely closed, narrowed, or replaced.
What to keep in mind
The paper explicitly says it is not a complete theory of prediction, language-model inference, learning, phenomenology, or runtime verification. The available abstract does not describe empirical testing or limitations beyond this scope statement.
Key points
- The paper argues that finite capacity makes the future operationally shaped by present traces, expectations, constraints, admissibility conditions, and witness availability.
- It introduces three terms: anticipatory trace, trace foreclosure, and horizon governance.
- The authors link the framework to AI generation, learning, TER runtime evidence, SFV translation, and the “No AI before trace” principle.
- Trace foreclosure is described as the premature closing, narrowing, or replacement of witnessable trace.
- The paper says it is a bridge note, not a complete theory of prediction, language-model inference, learning, phenomenology, or runtime verification.
Disclosure
- Research title:
- Finite-capacity futures are shaped by present traces
- Authors:
- Panagiotis Kalomoirakis
- Institutions:
- Institut national de recherche en sciences et technologies du numérique
- Publication date:
- 2026-05-17
- OpenAlex record:
- View
Get the weekly research newsletter
Stay current with peer-reviewed research without reading academic papers — one filtered digest, every Friday.