Ethan here assumes we have THE correct cosmological paradigm.
That is purely an assumption without much empirical evidence to back it up and a growing list of problems with the LCDM model, like the increased shakiness of the inflationary scenario, inability to identify the dark matter (virtually everything) or the cause of dark energy acceleration, the horrendous vacuum energy density crisis, etc.
There is a more rational argument for an infinite eternal cosmos that has evolving subsystems in “local” contexts, but in the global context it remains an eternal recycling hierarchy that is infinite and fully unified at all scales. Humans come and go. So do individual stars and galaxies, and even the whole observable Hubble bubble. However, the entire cosmos need not be equally mutable.
The cosmos is probably not going fade away, but bad physics hopefully will.