Validity of semiclassical bulk intuition and AdS/CFT at exponentially late times

Ascertain whether semiclassical bulk intuition and even the AdS/CFT correspondence remain valid at timescales t ~ exp(O(S_BU)), given that boundary time evolution induces only exponentially suppressed dynamics on baby-universe degrees of freedom in the random-encoding picture.

Background

In the authors’ model, logical baby-universe degrees of freedom are not reconstructable on either boundary, implying that boundary time evolution does not generate ordinary dynamics within the baby-universe sector except on exponentially long timescales set by the baby-universe entropy.

They emphasize that beyond such times it is uncertain whether usual semiclassical or even holographic reasoning should apply, highlighting a regime where the standard correspondence may break down.

References

Beyond such times, it is unclear whether semiclassical bulk intuition, or even the AdS/CFT correspondence itself, should be expected to apply.

— Baby universe as logical qubits: information recovery in random encoding  (2511.20747 - Mori et al., 25 Nov 2025) in Section 5 (Bulk picture: cloning puzzle and singularity)

Does this lesson carry over to more ``realistic'' settings? In particular, we would like to see if these lessons are also applicable at larger values of $N$ and for a generic potential.

— A pedagogical introduction to invariant theory and finite-$N$ holography  (2608.22952 - Koch et al., 24 Aug 2026) in Section 'Semiclassical Sectors'