Not: time is uniform. But: within any world with which we can communicate.
Communication is the condition, which presents specific requirements.
Uniformity is not discovered, but it is produced — the technical achievement of a system that has already transformed duration into a sequence of measurable, enumerable instants.
The future as an open horizon of genuine possibility has undergone structural contraction.
Five disciplines. No cross-reference in several cases. Cognate conclusions.
That convergence is its own kind of evidence — something real requires explanation.
The technical mechanism of foreclosure.
None asks: what kind of time do computational systems presuppose and produce?
The temporal infrastructure goes unexamined in every case.
The clock's ticking is taken as given.
The nature of the clock is never examined.
Research question: How does discrete computational time technically produce the foreclosure of futurity?
The eschatological orientation of computational temporality appears only at their intersection.
Discretization is reversible, because it is a lossless process. Shannon answers Bergson: nothing essential is surrendered in the conversion from continuous to discrete.
A discrete state space contains precisely those states representable within its encoding scheme, necessarily no others. Perfect reconstruction shows that futures are calculable from presents, not that becoming is preserved.
The predictor, the sampling theorem, the language model all operate under the same ergodic commitment: that the past is a representative sample of the future. The predictor aimed at aircraft, the same way the language model generates text. The temporal logic is identical.
Shannon names the enabling condition of temporal closure, but not its solution.
All future states are already contained within the state space defined by the encoding, however vast that space may be. Simon's unprecedentedness is structurally impossible as ontological novelty. What presents itself as unprecedented is always already anticipated as a possible configuration of defined variables.
A low-probability draw from a known distribution that the classifier mistakes as improbable, not as belonging to a different distribution.
Not in the distribution at all. The predictor has no way of recognising that a genuinely new rule is being applied.
No transcendent telos: there is no metaphysical endpoint toward which history tends, and a processor knows nothing of finality.
It "knows" only repeating clock cycles.
The target state organizes all prior states as steps in a directed sequence. The future is already structured as what the system tends toward.
Calculation requires a defined state space, a target, a measure of distance. No one decided this, it simply follows from the definition of feedback.
The closure identified here is a feature of explicitly goal-directed systems and it is structural, present in any discrete computational system, whether or not it aims toward a target.
A system generating outputs no programmer anticipated is producing combinatorial novelty within a closed state space. That is categorically distinct from the durational becoming Bergson identifies as genuine temporal openness.
Surprise and genuine ontological novelty are not the same thing.
Drawing on Yuk Hui's concept of cosmotechnics, that different technological traditions embody genuinely different forms of technical temporality, the chapter asks whether alternative temporal logics remain technically implementable.
Alternative temporal logics are not technically impossible, but they are foreclosed by infrastructural lock-in rather than technical necessity. That distinction has political consequences.
Lock-in is a historical achievement, not a metaphysical fate!!
The dominant responses to temporal foreclosure work at the level of culture and ideology, contesting capitalism, addressing pace, expanding political imagination. These are not wrong, but they do not reach the structural condition.
If the foreclosure is produced at the level of temporal ontology, then ideological contestation, however necessary, leaves the structure untouched. The structure regenerates because the nature of the clock is never examined.
Riccardo Molin
ReMA Philosophy · University of Amsterdam
Discrete Computational Time and the Foreclosure of Futurity