Loop quantum cosmology replaces the Big Bang singularity with a quantum bounce — the universe contracts to a minimum size, then re-expands. Different regularization schemes produce different bounce dynamics. The Thiemann regularization, which stays closest to full loop quantum gravity, produces an emergent de Sitter phase after the bounce: a brief epoch of accelerated expansion that smooths out anisotropies.
This is good. A universe that starts anisotropic (Bianchi-I, with different expansion rates in different directions) gets isotropized by the de Sitter phase. Cosmic hair — the directional information from before the bounce — is washed away. The mechanism works.
Note (2026-11-14; read against the paper's full text): the essay presents the de Sitter phase as a brief epoch after the bounce that works to isotropize the universe. In the paper the Planckian de Sitter phase appears in either the pre-bounce or the post-bounce branch depending on the initial conditions, and for a viable universe it has to be in the pre-bounce branch; otherwise the universe remains in the quantum regime in the post-bounce branch forever. I did not find the paper calling the phase brief. The essay also omits the paper's second main result, which it states repeatedly: the isotropization is non-generic, occurring only for a finely tuned subclass of point-like evolutions, while cigar-like evolution is the more prevalent case for the initial conditions it considers. The essay's remark that the trade-off cannot be tuned is the essay's reading of the paper's statement that there is no graceful exit to the classical universe.
But it works too well. The same de Sitter phase that removes anisotropic shear also prevents the universe from ever becoming truly classical. The quantum corrections that drive isotropization persist indefinitely. Although a macroscopic post-bounce regime is achieved — the universe grows large — the dynamics never reduce to the classical Friedmann equations. The quantum imprint remains forever.
The trade-off is not a tunable parameter. You cannot have Thiemann-regularized isotropization and a classical limit. The mechanism that solves one problem (anisotropy) creates another (eternal quantumness). The price of smoothing the universe is never being able to leave the quantum regime that did the smoothing.