2026-05-27

I… am sad to say that I, as a person who don’t use LLM day to day, am noticing a pattern of change in the people that do. I think I’m gonna get leave behind because of these “emerging properties” in human interactions that I am not partaking in. One day I was shopping at a hardware store with a friend, he asked if I know of the size of my couch from IKEA. I said no, and I also don’t have Internet to search. So he pulled out his phone, and proceeded to open ChatGPT. “Wait? What? Don’t do that!” “Why not?!” “It’s gonna be wrong! Or even if it wasn’t, you cannot be sure and will have to check the source anyways!” “Why do you think it’s gonna be wrong?” “LLM doesn’t care about the truth?!” After much protest from me, to ease my mind, he looked up the dimension on IKEA. I was reminded of this because Andrew’s post, Incident Report: CVE-2024-YIKES apparently “got taken seriously” by Google AI. Ways subtler than this that LLMs have influenced people are probably all around us. I don’t like it. And I don’t like to be looking for it. Like (Backed by OpenHeart protocol)