Comments on the black hole information problem

TL;DR

Problem: how to reconcile unitarity with semiclassical EFT in black hole spacetimes. Approach: argue for nonlocal effects suppressed by $e^{-S_{bh}}$ arising when restricting to time slices with small back-reaction, within an AdS/CFT unitary framework. Findings: outside observers can retrieve all information via Hawking radiation while inside observers experience effectively local physics until near the singularity; the nonlocality is nonperturbative in $1/N$ and constrained by a boost bound. Significance: provides a concrete mechanism for information retrieval compatible with locality for infalling observers and offers a framework connecting EFT, holography, and black hole evaporation.

Abstract

String theory provides numerous examples of duality between gravitational theories and unitary gauge theories. To resolve the black hole information paradox in this setting, it is necessary to better understand how unitarity is implemented on the gravity side. We argue that unitarity is restored by nonlocal effects whose initial magnitude is suppressed by the exponential of the Bekenstein-Hawking entropy. Time-slicings for which effective field theory is valid are obtained by demanding the mutual back-reaction of quanta be small. The resulting bounds imply that nonlocal effects do not lead to observable violations of causality or conflict with the equivalence principle for infalling observers, yet implement information retrieval for observers who stay outside the black hole.

Figures (3)

• Figure 1: Timeslices can be constructed in the semiclassical geometry that simultaneously intersect most of the Hawking radiation and a freely falling observer who crosses the horizon early on.
• Figure 2: A set of timeslices satisfying the back-reaction bound, and for which an observer hovering close to the horizon remains at low energy.
• Figure 3: Signal propagation from ${\cal O}_{free}$ to ${\cal O}_{accel}$ is restricted by the black hole geometry.