In heavy ion collisions, a nearly perfect fluid, known as the quark-gluon plasma, is created. Selecting the collisions with the largest final state multiplicities, it is predicted that the average transverse momentum rises as a function of the final state multiplicity. Under simplifying assumptions, one can derive a precise formula for the slope of this rising trend, which is proportional to the squared speed of sound of the fluid at a particular temperature. This temperature depends on the collision energy, allowing one to in principle obtain the temperature dependence of the speed of sound by analyzing collisions at multiple collision energies.  In this talk we will discuss the details of a recent measurement by the CMS Collaboration of the mean transverse momentum of charged hadrons produced in ultracentral PbPb collisions as a function of the charged particle multiplicity.  The slope of the rising trend is observed and is compared to hydrodynamic simulations combined with lattice QCD predictions. However, it turns out that the simplifying assumptions used to derive the formula are perhaps not satisfied in real collisions at the LHC. While this does not invalidate the measurement, it does complicate its interpretation, requiring careful analysis to determine the relation of the measured slope and the speed of sound.