The more that our models (using only "dumb physics") are able to effectively model the universe, the less we should assume that the large scale architecture of the universe - even the "medium" scale at the level of stars - are affected by the evolution of intelligent life.
Another way of saying this is that physical simplicity apparently dominates the architecture of stars, galaxies, and superclusters, without the complexity that we see in things like genomes and nervous systems (and the complex behavior those systems allow). If we are able to differentiate something usefully called "life" from background noise, then this complexity is certainly a core feature. At the scale of the universe, with each additional non-puzzling observation we make, it seems more certain that life has not had much effect. When we see a few stars that we can't understand, like KIC 8462852 or Fomaulhaut, that might mean some living things have crawled out of their respective primordial soups for long enough to build Dyson spheres, and we should be happy. But when we see something like dark matter, that's so mysterious it requires a whole new subatomic particle, we should rejoice! Maybe THAT is where everybody is, and that's is the ultimate fate of intelligence, uploaded at the end of evolution into some kind of ether! The sky's the limit! (Until of course, we find out that dark matter is just boring, simple, basic dumb predictable stuff.)
Whether this means that intelligent life does not appear (often), does not last long enough to have an impact, or has impacts at spatially smaller levels than this (see involution), is another question.
To argue that we can't know the impact of intelligences alien and greater than our own is to argue that we shouldn't bother talking about it, because we can't tell if any one proposition about alien intelligence is more likely to be true then another. That's a classic PEP (pointless epistemological problem).