Honestly, many bad situations can improved simply by understanding the risks involved and having a decent idea of where a potential threat is.

Flipping the script a bit, I have been a home automation early adopter for a good number of years and have relied on voice assistants, cloud accounts and all that garbage. Now that local AI is better developed, building everything I want in-house is feasible and I can stop giving my data away. I knew what my risks were and planned around it.

A person can err on the side of paranoia, or, they can simply not give a fuck. Both choices are valid and I use them both in my own weird ways.