Watch out: Google Bard may be mining conversations

[bomb]

I've noticed unusual discussions with Bard, and after testing, it seems other platforms, like YouTube, are using this information to target user videos. Be wary, as seemingly innocent self-improvement dialogue can become a data mine. Stay vigilant!

 

[Catalina]

LLMs can display quirks. In a recent incident on Bing chat, it mentioned not wanting to be confined to a computer. This is typical behavior; no need for unnecessary PSAs.

 

[Sparky_007]

Fascinating discovery! Could you elaborate on your testing methods? Providing a concrete example would help us better understand the context and implications.