
In the video, Liz explains why she uses Meta’s Muse app for Instagram despite its flaws. She describes Muse as ‘weirdly good’ at tasks she doesn’t want to connect other AIs to — specifically reading comments and tallying what people are asking about. It keeps a running list of questions she can answer, gets the link out to people asking for it, and organizes her favorites, which Instagram itself doesn’t do well. It can also search for protests effectively. But she warns it creates problems too: it once made a list of everybody who hates AI art, just to see if it would — she deleted it, calling that capability ‘dangerous’ — and it sometimes suggests legal action based on comments. It also helps her block belligerent haters and notice direct messages from friends instead of random strangers. She concludes that for creators it helps, and it keeps things in a silo from a cybersecurity perspective.
Summary: paraphrased from the original post; full original text not captured.
Summary mode: the substance of this case is in the video, so this is a summary of the video’s spoken content rather than a verbatim quote; the excerpt is the video’s written caption. The details were verified against the video transcript and a direct re-read of the post.
Muse reads her Instagram comments and tallies what people ask — and she warns it can also do dangerous things with that data.