End-of-the-World Branes and Inflationary Predictions for Rocky and Swampy Landscapes

TL;DR

The paper develops a local Wheeler–DeWitt measure for a multiverse of quasi-de Sitter vacua, linking the finite-dimensionality of dS Hilbert spaces (via the Cosmological Central Dogma) to explicit anthropic predictions. Predictions hinge on a rate equation that combines vacuum creation from nothing with tunneling between dS vacua, making the inflation scale highly sensitive to creation mechanisms and the presence of end-of-the-world (ETW) branes. Across swampy (no metastable dS vacua) and rocky (many dS vacua) landscapes, Hartle–Hawking creation tends to bias toward low inflation scales and is observationally disfavored, while Linde–Vilenkin-type creation with nontrivial topology or ETW-brane–driven processes can shift predictions depending on topology and brane tensions. The work highlights strong links between UV quantum gravity data (brane dynamics, topology) and macroscopic cosmological observables (the inflation scale and primordial gravitational waves), offering avenues to test string-inspired multiverse scenarios with upcoming CMB experiments.

Abstract

Making cosmological predictions in a multiverse is a fundamental theoretical challenge. Assuming that (quasi-)de Sitter vacua are quantum mechanically described by a finite-dimensional Hilbert space, we develop a detailed framework for making explicit anthropic predictions. A key challenge which we attempt to overcome arises because, almost unavoidably, cosmologies that asymptote to Minkowski space exist. We then apply our framework to predicting the scale of inflation. We find that, even if eternal inflation is allowed, our predictions depend on creation rates of universes from nothing. These, in turn, are highly sensitive to the existence of end-of-the-world branes. The rates for the creation of universes from nothing are the dominant ingredient for `Swampy Landscapes', which may have no metastable de Sitter vacua but only slow-roll solutions. In `Rocky Landscapes', where long-lived de Sitter vacua are abundant, tunneling rates between such vacua represent a further key factor for deriving predictions.

Figures (5)

• Figure 1: Creation and tunneling for a toy landscape consisting of three dS, one supersymmetric Minkowski (M), one non-supersymmetric and one supersymmetric AdS vacuum ('AdS$_4$' and 'AdS$_5$' respectively). Note that the labels distinguish different vacua and have nothing to do with dimensionality, which is always 4. To avoid overburdening notation, we have not put labels on the arrows for decays to Minkowski and AdS vacua.
• Figure 2: Approaches to the Minkowski-space infinity problem.
• Figure 3: An AdS bubble with crunching surface (wiggly line) and reheating surface (red line) is shown.
• Figure 4: Illustration of a Minkowski bubble inside a dS bubble (left) and a colliding bubble system (right). On the left, the part of the reheating surface that can send signals into the bubble is finite, meanwhile on the right it is infinite.
• Figure 5: The creation of a spherical universe with a subsequent formation of a dS bubble (left) and Minkowski bubble (right) are displayed. The red line denotes the bubble wall and the blue line illustrates the location of the reheating surface. The diagonal dashed line on the left shows the cosmological horizon for an observer located on the vertical solid black line on the far left.