This is the entry point: one hypothesis, four claims, its limits, and what would refute it. The full introduction is seven to eight times longer — a link leads there from here.
In one sentence: in an unstable state, intersections of fractal objects arise; their projections appear as singularities in lower dimensions; and where these turn back into themselves, they iterate.
Physics treats the speed of light, the Planck length, the uncertainty principle and the Bekenstein bound as separate limits, each with its own derivation. ISI asks whether they are projections of one and the same boundary, seen from different dimensions. This is a question rather than a claim, and the framework exists to make the question measurable.
Picture a three-dimensional object casting shadows on two different walls: a circle on one, a square on the other. Anyone who sees only the walls describes two things and finds two rules. ISI suspects the same about physical limits: not separate laws, but projections of one structure, differing by the dimension you look from. The analogy breaks at one point, worth saying plainly: what projects here is not a single object but a network of intersections, and the projection is itself part of the phenomenon.
The framework is not made of equally strong sentences. These four are the core, each with its epistemic status.
A detectable particle is identical to the intersection of singular sub-manifolds. This is the first of the Core axioms: a definition the rest builds on, not a measurement.
A particle is not simply in or out of a dimension: its participation takes fractional values. It follows that the photon moves at c because mass participation is absent from it.
Two impossibilities stand side by side: stopping is impossible, arriving is impossible. What remains is iteration. The K5 theorem proves the second: convergence never completes in finitely many steps.
ISI has no independent time dimension. This is the framework reading rather than an experimental result — experiments (Barontini, zeptosecond, top-quark) are compatible with it but do not prove it.
This section matters as much as the previous one, and in longer material it usually gets lost.
A hypothesis is worth the risk it takes. ISI keeps fourteen numbered predictions; three appear here, all with a refutation route available to current technology.
If dark matter is confirmed as WIMP annihilation — with a density-squared signal and a sharp mass cut-off — the ISI reading of a collective, continuous spectrum fails. The Fermi 20 GeV gamma halo analysis is a live test.
Attosecond and zeptosecond metrology measures the residual beyond QED. If that residual is absent, or does not scale as predicted, the density claim about fractal space falls.
Self-replicating systems should not work with fewer than four symbols. A stable two-symbol self-replicator would overturn the theorem.
One result is registered in advance as a consistency test: exceeding the Tsirelson bound at any energy would remove the geometric basis of the framework.
Three directions open from here, depending on whether you want to understand, to verify, or to read the source.
Hypothesis-generating framework, prior to peer review. Formal definitions, derivation targets, interpretations and testable predictions are marked separately: where the page offers an interpretation, that is not evidence, and where it offers a prediction, what would falsify it is stated. The current working version is ISI Core v5.0; claims withdrawn from earlier versions are recorded in the change log and marked in the affected sections.