I know people who takes hours in comfyUI making a workflow, tweaking aspects, choosing different nodes, adding several layers of different diffusion models.
You can use an AI generator just by making a prompt "make me a pretty giraffe" same I can take my phone a snap a quick picture. But same as a professional photographer can take hours chosing composition, camera configuration, then tweaking the result.. a person who want to make a good AI image can take hours or days improving and tweaking the workflow.
For instance, this is a workflow example, a easy one, not even the most complex I've seen:

That could take a long time to make, because the person had a specific vision on what they want the tool to produce, and can really steer it into producing exactly what they want.
I think a lot of hate, as always, come mostly from ignorance. Once you know the time and effort that someone can put into this, it's harder to discredit them.