In my experience it doesn’t take long until you use such complex offerings from the cloud vendors, you need those ops engineers anyways. Just with slightly different skillsets.

I'd say you need people with certain skill sets anyways but at a certain scale you have to get specialized people for some service. Database admins, kubernetes admins, network admins. At a small scale that can be one or two people. But if you want 24/7 with a bigger scale you need multiple people for each role. You have to find them, pay them, schedule their absences.

To some management types it looks like a good deal to not deal with that and just let Amazon/Microsoft/Google/etc. deal with finding people to support the service and just pay a bit extra to the infra cost. Then you can only hire cloud infra admins. I don't think it works that way but that is what I have observed.