A tabletop experiment at the University of Birmingham has put 24,000 ultracold atoms to work as a miniature universe with no external clock to read. Physicist Giovanni Barontini cooled a cloud of rubidium atoms to within a few billionths of a degree of absolute zero, then tested a question that quantum gravity has left open for decades: can time be defined from inside a closed system? His published answer, from June 2026, was yes.
AI, explained every day: The Deep View is the go-to daily newsletter for 600,000+ founders, knowledge workers and professionals who want to stay up to date with artificial intelligence.
We may earn a commission if you sign up.
Barontini, a professor at Birmingham, framed the puzzle in the university’s release by noting that in some theories of the universe, especially quantum gravity, time does not appear as a built-in feature. The experiment tries to show what time could be instead, using only the changing contents of a sealed system as the measure of before and after, with no laboratory timer standing in for the missing one.
Wheeler-DeWitt and the missing time variable
The theoretical backdrop is the Wheeler-DeWitt equation, a quantum-gravity formula that describes the whole universe without any external time variable in it. Nothing in the equation ticks. If the universe is everything, there is no outside clock to consult, and the familiar sense of before and after has to arise from relationships among the contents of the universe itself. Physicists have proposed that entropy, a measure of how a system’s particles are distributed, could supply that ordering. Until this experiment, the idea lived mostly on paper. The ScienceDaily summary of the Birmingham work describes the apparatus as an isolated quantum system built to put that proposal under measurement.
Everyday physics hides the problem because a laboratory always has an outside observer holding a watch. A closed universe has no such observer. Entropy offers a way out because, in an isolated system, it changes in one preferred direction, and that change can be used to label moments in order. A model built on this reading has to show three things: that the ordering is consistent, that it has a direction, and that ordinary quantum dynamics survives when written in the new variable. Barontini set out to test each of those against real atoms rather than a computer simulation.
Bright and dark regions in a laser-cut atom cloud
Two laser beams tuned to different frequencies formed a thin barrier through the middle of the cloud. One side became the bright region, which was observed, and the other the dark region, which was not. Atoms could cross the barrier, so the observed side kept gaining and losing particles. As that happened the bright region repeatedly expanded and contracted, a loose analogy for a Big Bang followed by a Big Crunch. The release is explicit that this is a simplified quantum model and not a re-creation of cosmic events. The paper, “Testing the problem of time with cold atoms,” appeared in Physical Review Research.
About 24,000 atoms made up the system, a count large enough for the particle distribution on the bright side to be followed as it shifted. The EurekAlert copy of the release identifies them as rubidium, held near absolute zero, the temperature range where laser cooling and trapping removes the random thermal motion that would otherwise swamp a delicate quantum measurement. Birmingham’s account does not describe the apparatus in further detail, so questions about trap geometry and imaging stay with the paper itself.
Entropic time and the Schrödinger equation
With no clock in the apparatus, Barontini let entropy do the timekeeping and calls the result entropic time. Whenever the particle distribution in the bright region changed, time advanced; when the change stopped, time effectively stopped with it. Three properties stand out in the University of Birmingham release: entropic time runs in a single direction, which gives the arrow of time; it orders events correctly through the expansion and contraction cycles; and its rate speeds up or slows down depending on how fast entropy changes. A version of the Schrödinger equation can be written in terms of entropic time, so the standard machinery of quantum mechanics still makes predictions with it.
The scheme works because the observed region is the only part of the apparatus that leaves a clock-like record. Counting how many atoms sit in the bright region at each moment gives a distribution, and the entropy of that distribution is the quantity that moves. When atoms flow in and out quickly, entropic time races ahead. When the flow stalls, the measure goes quiet, which is the sense in which the work reports time halting while the atomic distribution holds steady.
Barontini presented the result as a first. His statement said the study provides the first controlled experimental evidence that time can be defined by changes within a system, rather than as the external ticking clock most people picture. The Debrief, which covered the paper after publication, relayed the same claim and noted that when the atomic distribution held steady, time appeared to halt.
Limits of a lab-scale test of time
The experiment does not show that time is an illusion, and it does not show that the real universe lacks a fundamental time. What it offers is a controllable platform where a competing definition of time can be checked against measurements. Birmingham’s release adds that the approach could be extended to simulate black holes or to test other proposals for how time emerges, which would bring more of the quantum-gravity literature within reach of laboratory comparison.
Questions of scale remain open. The evidence comes from a single cloud with one barrier geometry, and the summary names no collaborators or funders. The journal page for the paper holds the full analysis. Whether other laboratories reproduce entropic time with different atoms or barriers will decide how far the claim travels, and whether entropy-based clocks survive in systems with many more particles, or with the gravitational effects that motivate the whole exercise, is a separate matter that this experiment leaves untouched.
This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.
AI, explained every day: Get The Deep View →
We may earn a commission if you sign up.
More from Morning Overview
