Time to start using an LLM to farm out massive amounts of faked computer-generated "human-generated" data about myself, then.
I've actually thought about something similar more than once. I think it would be interesting to have multiple accounts across multiple websites, all making posts and interacting with content across the algorithmic spectrums. That way, when mass-data harvesting via LLMs begins to become much more common, the data about what I think or believe is so tainted that it would be impossible to decipher a singular sincere ideology.
LLMs make practically any human-generated data valuable. These will be collected as long as it’s legal, to a greater degree now than before.