No telescope has spotted another universe, and no experiment has confirmed that one exists beyond our own. But “multiple universes” is not a single wild idea; it is a label covering several distinct theoretical proposals, some of which fall out naturally from physics we already trust. The question is less “did someone dream this up?” and more “how many roads in well-established physics seem to point in that direction?” The answer, perhaps surprisingly, is several, and they differ so dramatically in what they mean by “another universe” that lumping them together can obscure what is actually being debated.
The Simplest Version: Space Just Keeps Going
The most straightforward multiverse idea requires no exotic physics at all. If the universe is spatially infinite, and if the laws of physics are the same everywhere, then the sheer volume of space guarantees that every possible arrangement of matter repeats somewhere. Inflationary cosmology, the leading framework for the universe’s earliest moments, generically predicts an infinite space filled with regions that have gone through the same kind of Big Bang expansion we did. Each of these regions, sometimes called Hubble volumes, realizes all possible initial conditions. In a framework proposed by physicist Max Tegmark, this is called the Level I multiverse, and it predicts that an identical copy of you exists roughly 10 to the power of 10 to the 29th meters away.
1Cambridge University Press. The Multiverse HierarchyThis version of “other universes” is really just other patches of the same universe, too far away for light to have reached us since the Big Bang. There is nothing mysterious about the physics. The strangeness comes only from the implication: if space is truly infinite, probability alone means every possible configuration of atoms, including ones arranged exactly like the room you are sitting in, occurs somewhere. You could never visit these other patches or communicate with them, because the expansion of space carries them away faster than light. But they are, in principle, out there, governed by the same laws and composed of the same particles.
Eternal Inflation and Bubble Universes
A more dramatic picture emerges from a feature of inflation itself. In many versions of inflationary theory, the process that drove our early universe’s rapid expansion does not stop everywhere at once. Instead, inflation ends locally in patches, each of which “reheats” into a hot, expanding region that looks like its own Big Bang. Meanwhile, inflation continues elsewhere and keeps going forever, continually spawning new Big-Bang-like regions.
2arXiv. Eternal Inflation, past and futureEach of these regions is a bubble universe. Unlike the Level I picture, where every region shares the same fundamental physics, bubble universes can in principle have different physical constants, different particle masses, even different numbers of spatial dimensions, depending on how the underlying fields settle as inflation ends in each bubble. This is Tegmark’s Level II multiverse, and it is far stranger than the Level I version because the “other universes” are not just copies of ours but genuinely alien in their basic properties.
Eternal inflation is not a fringe add-on to standard cosmology. It arises naturally in many of the inflationary models that best fit observational data. The controversial part is not whether eternal inflation happens, if standard inflation is correct, but whether the bubble universes it implies should count as real predictions or as untestable side effects of the math.
Quantum Branching and the Many-Worlds Interpretation
An entirely separate route to multiple universes comes from quantum mechanics. The Many-Worlds Interpretation, proposed by Hugh Everett in 1957, takes the mathematics of quantum theory at face value: when a quantum system can produce more than one outcome, every outcome actually occurs, each in its own branch of reality. There is no mysterious “collapse” that selects one result. Instead, the universe splits, and each branch proceeds as if its outcome were the only one.
This idea has had a complicated reputation. For decades it was seen as philosophically extravagant, but it has gained significant traction among physicists who find the alternatives worse. Physicist David Deutsch argued in the 1980s that the computational power of quantum computers places “an intolerable strain on all interpretations of quantum theory other than Everett’s,” since the calculations seem to require parallel processing across branches that, in his view, must be physically real.
3Proceedings of the Royal Society of London. A. Mathematical and Physical Sciences. Quantum theory, the Church–Turing principle and the universal quantum computerModern work on Many-Worlds focuses on making the branching structure precise. How do you count branches? How do probabilities emerge in a theory where everything happens? Recent research has proposed new branch-counting rules grounded in decoherence theory, the physics of how quantum systems lose their quantum character through interaction with their environment. These rules aim to put the branching on firmer mathematical footing, though the debate is ongoing.
4Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences. Branch-counting in the Everett interpretation of quantum mechanicsIt is worth noting that quantum branching produces “universes” of a fundamentally different kind than eternal inflation does. These branches are not distant regions of space you could fly to. They are alternative histories occupying the same space but no longer interacting with each other after decoherence takes hold. The word “multiverse” covers both ideas, but the underlying mechanisms have almost nothing in common.
The String Theory Landscape
String theory adds another dimension to the conversation. In the early 2000s, theorists discovered that string theory does not predict a single set of physical laws. Instead, it contains an enormous number of possible vacuum states, each corresponding to a different way the theory’s extra spatial dimensions can be curled up. These vacua can have different particle spectra, different force strengths, and different values of the cosmological constant. Many of them turn out to have metastable local minima with a small cosmological constant, meaning they could describe a universe that looks roughly like ours.
5Universe. The String Theory LandscapeThe number of these vacuum configurations is staggering, commonly estimated at something like 10 to the 500th power. Combined with eternal inflation, the landscape provides a mechanism: different bubble universes settle into different vacua, each one realizing a different set of physical laws from the vast menu string theory provides. Our universe, with its particular particle masses and force strengths, would then be one bubble among countless others, each with its own version of physics.
Critics point out that a theory predicting 10 to the 500th possible outcomes without telling you which one corresponds to reality is, by some standards, predicting nothing at all. Supporters counter that this is the wrong way to think about it: the landscape, combined with anthropic reasoning, might explain why our universe has the properties it does without requiring a unique prediction.
Cyclic Universes and Colliding Branes
Not every multiverse idea involves parallel realities existing simultaneously. In cyclic cosmology, the universe goes through repeated cycles of expansion and contraction. The ekpyrotic model, developed in the early 2000s, describes the Big Bang not as the beginning of time but as a collision between higher-dimensional surfaces called branes. Before the Bang, the universe slowly contracts through an ekpyrotic phase, which resolves the standard puzzles of cosmology and generates the nearly uniform distribution of matter we observe.
6Physics Reports. Ekpyrotic and cyclic cosmologyIn these models, dark energy is reinterpreted as a weak attractive force between our brane and a neighboring one. That force eventually draws the branes together again, triggering a new ekpyrotic phase and a new collision, starting the cycle over. Each cycle is, in a sense, its own universe, a complete cosmic history from bang to crunch to bang. The “multiple universes” here are separated by time rather than by space or quantum branching.
One testable distinction: cyclic models predict gravitational waves with a blue spectrum, meaning higher-frequency waves are stronger. Standard inflation predicts a red spectrum. If future gravitational-wave observatories can measure the primordial spectrum at the relevant frequencies, this would be a genuine way to distinguish between the two frameworks.
Cosmological Natural Selection
Physicist Lee Smolin proposed a multiverse mechanism that borrows logic from biology. The idea starts with a speculative but physically motivated assumption: when a black hole forms and a singularity develops, quantum gravity effects might cause that singularity to bounce, spawning a new expanding region of spacetime, essentially a baby universe. Following a suggestion originally due to John Wheeler, the fundamental constants in each new region would differ slightly from the parent universe, like random mutations.
7arXiv. The fate of black hole singularities and the parameters of the standard models of particle physics and cosmologyOver many generations of universes, this process naturally selects for physical constants that maximize the production of black holes, because universes that make more black holes produce more offspring. Smolin pointed out that this is a testable prediction: if the theory is correct, our universe’s parameters should be near-optimal for black-hole production, and small changes to any fundamental constant should reduce the number of black holes formed. That makes it one of the few multiverse proposals that sticks its neck out with a concrete, falsifiable claim, though the evidence remains inconclusive.
What the Data Actually Shows
Can any of these ideas be tested? More than you might expect, though none has produced a smoking gun. One of the more intriguing observational leads involves the CMB Cold Spot, an unusually cold region in the cosmic microwave background about 5 degrees across. One proposed explanation is that it is the imprint of a collision between our universe and another bubble universe during an early inflationary phase.
8Monthly Notices of the Royal Astronomical Society. Evidence against a supervoid causing the CMB Cold SpotThe competing explanation, that the Cold Spot is caused by a large void in the distribution of galaxies along that line of sight, has been weakened by recent observations. A 2023 analysis using gravitational lensing data ruled out the supervoid interpretation at high confidence, leaving the Cold Spot’s origin genuinely open.
9Journal of Cosmology and Astroparticle Physics. The CMB cold spot under the lens: ruling out a supervoid interpretationThat does not mean the bubble-collision explanation wins by default; other mundane explanations remain possible. But theorists have been developing increasingly precise predictions for what a bubble collision would look like in the CMB. Recent work has produced the first analytic expressions for specific observable signatures of such collisions, finding that next-generation experiments like CMB-S4, combined with galaxy surveys, could improve our ability to detect or rule out these signals by roughly an order of magnitude over what current CMB data alone can do.
10arXiv. Eternal inflation bubble collision signature on CMB remote dipole and quadrupole fieldsAnother approach looks for signs that the universe has a finite, repeating geometry, which would show up as matching patterns in the cosmic microwave background. The Planck satellite searched for these patterns and found none. Its analysis constrained any repeating structure to be larger than about 94% of the distance to the last-scattering surface at 99% confidence, meaning if the universe does repeat, it does so on scales too large for us to see.
11Astronomy & Astrophysics. Planck 2015 results. XVIII. Background geometry and topology of the UniverseThe Testability Debate
The deepest objection to multiverse theories is philosophical rather than empirical: if you cannot observe the other universes, does proposing them count as science? This question has split physicists more sharply than almost any issue in modern cosmology. The multiverse hypothesis seems to conflict with the methodological standards traditionally accepted in physics. It is not generally falsifiable in the classic sense that Karl Popper described, which has prompted some physicists to argue that the standards themselves need updating for a cosmos that may simply be too large and too varied to fit neatly into the framework of laboratory-scale hypothesis testing.
12Studies in History and Philosophy of Science Part B: Studies in History and Philosophy of Modern Physics. Testability and epistemic shifts in modern cosmologyThe defenders have a point worth considering: we do not reject atoms because no one has seen one with the naked eye. We accept their existence because the theory that includes them makes predictions we can check. If eternal inflation makes testable predictions about the CMB, and those predictions pan out, the other bubble universes that the same theory implies gain credibility even if we cannot visit them. The logic is indirect but not unprecedented in physics. What critics rightly worry about is the gap between “the theory predicts other universes as a side effect” and “we have verified the specific predictions that distinguish this theory from alternatives.” That gap remains wide.
Bridges, Wormholes, and Connections Between Theories
Some of the most intriguing recent theoretical work blurs the lines between different multiverse ideas. The “ER = EPR” conjecture, proposed by Juan Maldacena and Leonard Susskind, suggests that Einstein-Rosen bridges (wormholes in spacetime) and quantum entanglement (the spooky connection between distant particles) are actually two descriptions of the same underlying phenomenon.
13Universe. The “ER = EPR” Conjecture and Generic Gravitational Properties: A Universal Topological Linking Model of the Correspondence between Tripartite Entanglement and Planck-Scale WormholesIf this conjecture holds up, it would mean that the quantum connections linking entangled particles are, in some deep sense, tiny spacetime tunnels. This does not directly prove a multiverse, but it hints at a picture in which the fabric of spacetime and the fabric of quantum theory are woven from the same thread. That kind of unification could eventually clarify whether the “branches” of Many-Worlds and the “bubbles” of eternal inflation are truly separate ideas or different faces of one underlying structure. For now, though, this remains firmly in the territory of theoretical exploration.
Why Not Just Simulate the Whole Thing?
A tangentially related question that captures public imagination is whether our universe might itself be a simulation running in some grander reality. A 2025 analysis took this question seriously as a physics problem and asked how much energy or computing power it would take. The answer was discouraging for simulation enthusiasts: simulating even just Earth at a resolution consistent with the high-energy neutrino observations we already have would require amounts of energy that are, in the paper’s word, “astronomically” large. Simulating the entire visible universe is far worse. The authors concluded that it is simply impossible for a universe sharing our physical properties to simulate our universe, regardless of future technology.
14Frontiers in Physics. Astrophysical constraints on the simulation hypothesis for this Universe: why it is (nearly) impossible that we live in a simulationThe one escape hatch: a universe with very different physical properties, perhaps one where computation is far cheaper, might manage it. In a strange way, this loops back to the multiverse: the simulation hypothesis only works if there are universes with radically different physics from ours, which is exactly what some multiverse models predict. The argument does not prove anything either way, but it illustrates how these seemingly separate speculative ideas keep bumping into each other.