@straykat: Why would they? They're not biological. They exist "forever". They need only stop and "store" biologicals because of their programming. Because - let's be serious... if the Catalyst was capable of human thought do you think it would not have seen the terrible flaws in its programming after a few hundred cycles?
Honestly - I believe logic engines would come to the conclusions humans seem incapable of. The universe is vast and you (we, it, they, whatever) are not important as an individual. If anything about us is important - it is that we are participating in "life" - so is phytoplankton (and is actually far more important for life in general)
Policing "galaxies" is well beyond the boundaries of preposterous as far as stories goes... policing "the universe" would be a futile effort and I don't see any logic engine being stupid enough to think it's a worthwhile goal. We know - through the story - they ARE capable of policing the galaxy (regardless of how insane that is) - but even so, they're failing more and more each cycle.
The Catalyst should have known this inevitability was coming - and it probably would have - if it were anything more than a fancy program. As it was - the Catalyst was incapable of anything like original thought - synthesis likely strained the boundaries of its programming - a glitch it cobbled together (and also an ode to transhumanism)
Personally - I believe true AIs that could think for themselves - would tend toward fatalism. Only something programmed for "awe and wonder" would act like a human. Humans act that way because we are moved by WAY more than just logic. We're an irrational creature capable of moments of reason.
The one thing I applaud in their depiction of the Catalyst? How supremely incompetent it was over millions of years. Completely unwilling - or incapable - of any real significant change to its programming.