Just thinking about the way LLMs can analyze huge amounts of data to cross references and even use your writing style to dox you… well… seems like we gonna need some text spoofer tools that change each of your messages to a different grammar set and random typo generation so they all seem like sent by different users…
I think the latest outcry with LLMs unmasking pseudonymous users wasn’t due to their writing style and grammar. But they used the information and facts the posts contained. You’d basically need to switch up your writing style. But then also stop blogging about your favorite Star Trek series and Linux, and pick a new hobby to write about. And claim you’re a nurse from Uganda and random new things.
The Ugandan weather is so good today for my nursing career. Also I just adopted this squirrel.
This is why I rarely give information about myself, and when I do, I pair it with a bit of info that’s just a lie. So if people look through my posts/comments they’ll find conflicting information about my identity.