cross-posted from: https://lemmy.ca/post/31187638
Earlier today I came across a Reddit comment with a link to an Instagram post. The link had
?igsh=at the end.When I clicked on the link, I got this popup. It had a name and profile photo that was different from that of the post being shared.
Join Firstname Lastname on Instagram
See photos, videos, and more from Firstname Lastname.
[ Open Instagram ]
not now
I avoid link trackers. However, I did not realize it was this bad.
To my knowledge, TikTok does the same thing and lists the name of the person that shared the link. Assuming this increases engagement, any website could enable such a feature, even on old links that you shared in the past.
You should manually remove any trackers before sharing, or use an app for it.
Wait, you conduct yourself in a manner consistent with a reputation you’d like to maintain? Even when you don’t think anyone will find out who you are?
What kind of responsible shit is that? Don’t you know you’re on the Internet?
I guess I do think there’s a non-zero chance that somebody might find out. With advances in LLMs perhaps it will be possible to create a model that can recognise your internet posts by your writing style and/or certain facts or anecdotes you’ve mentioned more than once.
Hah, indeed. I was just playing off the John Wooden quote “The true test of a man’s character is what he does when no one is watching."
There’s another quote (I think from one of the writers at Forbes) that was modified to “… when everyone is watching”, which I think holds water, too.
Edit: apparently I was half right; https://www.forbes.com/sites/walterpavlo/2012/10/23/character-is-what-you-do-when-everyone-is-watching/
Fair enough, I’m not aware of that quote, but that makes sense. But yeah I do wonder if future employers will be able to find your online posts more easily, using LLMs. So it’s probably important to only say things that you can defend.
I’m more worried about false positives then an LLM actually being able to track me by writing style.
That’s a good point. But I still worry about possible genuine positives. For example, if i criticise a foreign government online because of things like human rights abuses, what if that government can use an LLM to identify me and other critics of them?
Maybe I’m just being paranoid though.