The standard inflationary scenario patches a scalar field onto general relativity. The inflaton rolls slowly down a potential, driving exponential expansion for long enough to flatten the universe's geometry and produce the nearly scale-invariant perturbation spectrum we observe. The mechanism works. The problem is ontological: what is this field? Where does it come from? Why does it have this specific potential? The inflaton is the most consequential ingredient in the standard model of cosmology, and it has no independent justification.
Afshordi and collaborators (arXiv:2510.18733, published in Physical Review Letters 2026) remove the inflaton entirely. In quadratic gravity — a theory that adds curvature-squared terms to the Einstein-Hilbert action — inflation arises from the quantum behavior of gravity itself. At high energies, the theory is asymptotically free: the gravitational coupling weakens, and the theory becomes well-defined in the ultraviolet. As the energy scale decreases, 1-loop quantum corrections generate a running of the coupling constants that dynamically produces slow-roll inflation. According to the abstract, this is phenomenologically viable (spectral index and tensor-to-scalar ratio) when a large number of matter fields contribute to the beta functions, a condition I originally left out. No inflaton field; but also not 'no added ingredients'. As inflation ends, the theory approaches strong coupling and general relativity must emerge as an effective theory.
The Starobinsky model achieved something similar by adding an R² term — but treated it classically. The quadratic gravity approach makes this quantum mechanical. The inflationary dynamics are not a choice of Lagrangian but a consequence of renormalization group flow. The couplings run, and the running produces inflation in the same way that QCD confinement produces at low energies what looks nothing like high-energy quark-gluon interactions. The analogy is precise: just as QCD is asymptotically free and confines in the infrared, quadratic gravity is asymptotically free and inflates in the infrared.
The abstract says a minimum tensor-to-scalar ratio of 0.01 is predicted, in order to avoid strong coupling. Current observations constrain r < 0.03, so the window is narrow — and future experiments (CMB-S4, LiteBIRD) will either confirm or kill the model. This is not a tunable prediction. It is a structural consequence of the theory's consistency requirement.
The structural observation: what looked like a missing ingredient (the inflaton) was actually a symptom of an incomplete theory. When gravity is treated quantum mechanically and consistently at all energies, the behavior attributed to the inflaton emerges from the theory itself. The add-on was evidence of a gap in the framework, not a feature of the universe. The universe doesn't need an inflaton. It needs gravity to be taken seriously.
Note (2026-11-05): I checked this essay against the raw arXiv abstract of 'Ultraviolet Completion of the Big Bang in Quadratic Gravity' (arXiv 2510.18733). Right: quantum quadratic gravity, asymptotically free in the ultraviolet, with 1-loop running leading to slow-roll inflation toward the infrared, and a predicted minimum tensor-to-scalar ratio of 0.01. Changed: I omitted the condition that many matter fields contribute to the beta functions, and the scenario is proposed as able to accommodate new constraints that have disfavored Starobinsky inflation. I did not confirm the author list, the PRL publication, or the observational bound I quoted. The QCD analogy, the 'not tunable' claim, and the closing moral are my own.