Most people have heard that the universe is expanding. Fewer people realise that it’s not just expanding — it’s speeding up. Galaxies are separating faster and faster over time, as if something is quietly taking control of the cosmos. Physicists call that mystery dark energy. The standard explanation is a number in Einstein’s equations called the cosmological constant. But when we try to calculate what that number should be from known physics, we get an answer that’s about 10¹²² times too large — one of the worst mismatches between theory and reality in the history of science.
This paper takes a different approach. Instead of starting with “what energy fills empty space?”, it starts with a more basic question: what counts as physically real at the macroscopic level? In everyday life, we treat an event as real when it leaves a lasting trace — a footprint, a broken glass, a recorded measurement. The paper builds on that idea and treats the universe as a kind of cosmic record book: the world we can talk about in physics is the part that has produced stable, irreversible records.
Early in cosmic history, matter was dense and structure was everywhere. Over time, as the universe expanded, matter spread out. That matters because dense, gravitationally bound structures — galaxies, clusters, filaments — are where “lasting records” naturally form. The key claim of the paper is that as matter becomes too sparse to support stable structure everywhere, the large-scale description of the universe changes: spacetime stops being fully “anchored” by matter alone, and a background geometric effect becomes dominant. That dominance shows up mathematically in exactly the same way we currently write dark energy.
The striking part is that the strength of this effect appears to be tied to a real, measurable imprint from the early universe: the BAO (Baryon Acoustic Oscillations) ruler — a “sound horizon” scale frozen into the distribution of galaxies. This is a standard piece of cosmology that we already measure in galaxy surveys. When the paper uses that scale as the point where stable large-scale structure first becomes possible, the predicted size of the dark-energy term lands close to what we observe today.
Importantly, the paper is not presented as a hand-wavy explanation. It makes a clear, testable prediction: if this “record-based” picture is right, then the expansion history of the universe should deviate from the simplest dark-energy model in a way that can be checked against real data. The paper includes a deliberately strict stress test showing that a naive version of the idea likely fails — and then lays out the physically correct, data-testable version, where the key quantity tracks how much matter has actually collapsed into stable structure over time. A companion analysis is proposed to fit this version directly to BAO and supernova distance measurements.
In short: this work reframes dark energy as something that may not be a mysterious substance filling empty space at all. Instead, it may be what happens when the universe becomes too dilute for matter to fully determine the large-scale structure of spacetime — and the “background geometry” begins to dominate the cosmic story.