Unconventional histories of the early universe
Pasquale D. Serpico (LAPTh, France)

Most of cosmology relies on a few measurements at some epochs, plus extrapolations using known physics.
Extrapolations are only safe if known physics (or some implicitly assumed properties of the unknown one) are valid.
Since we know (notably via some cosmological observations) that fundamental physics is incomplete, it is a reasonable expectation that some of these extrapolations may break down. I will illustrate these general considerations with two (relatively neglected) examples:
First, I will show how some features about cosmological events predicted within the Standard Model (the electroweak and QCD phase transitions) may be altered in a quite minimal BSM extension. Despite its simplicity, a dramatically altered history of the early universe may follow, even at scales naively expected to be under control, since “the Standard Model rules”. I will briefly outline some consequences for colliders as well as for cosmology.
Second, I will present CMB bounds on an “unusual” type of early universe relic, Primordial Black Holes, as a part or all of the dark matter.
In fact, conceptually their possible formation relies on the breakdown of the extrapolation of the smallness of primordial perturbations down to small scales.
Although they might be seen as a "nightmare" dark matter candidate, due to their purely gravitational interactions, I will show that this picture is too pessimistic, and that cosmological constraints and in particular CMB ones cover interesting parts of the parameter space.