Example - I have JSON based RPC so that systems from other vendors can talk to mine (enterprise backends written in C++). It works like a charm and has ben doing so for years. Here comes this architecture astronaut and tells ma that I should do GraphQL and proceeds to explain me how powerful and cool it is and how everybody and his cat uses it. So on the downside I will waste a gobbles of time and money, on upside - zilch because nobody gives a shit. The problem is already solved for us year ago so buzz off. And that guy could not give a single example of how it can help me. Just spreading FUD about existing things.

There are numerous opposite examples when I see that this new tech, lib, tool actually saves me time and money. I pay quite a few dollars for software tooling. Well if the tool does not offer perpetual license it is a no go for me then.

This is all that matters to me. On desktop for example I skipped moving to that .NET bandwagon and stayed with Delphi for my GUI desktop products. They worked 20 years ago and they work the same now. Single 10MB self updating exe with zero deployment issue. And free from numerous limitations imposed by UWP. Competitor is 1GB package with crapload of problems and every update turns a nightmare for customers. In my case all the time is spent creative stuff that brings me new customers / money instead of feeding someone else. Sure it costs me few hundred a year but that is peanuts.