When something (news) is, by nature, designed to inform you of the current state of reality, introducing any process that hallucinates (read: makes shit up) should not be anywhere near them.
People are free to use LLM plugins to “de-clickbait” headlines as they want, but keep that out of the news process itself.
Yeah, little control over presented articles was what led me to drop it when I was considering reddit alternatives but I’d kill for that de-clickbaiting feature to make way to RSS readers and link aggregators.
It’s not all bad. Artifact uses “AI” to change clickbait titles into more descriptive ones for example.
Still a slippery-slope, IMO.
When something (news) is, by nature, designed to inform you of the current state of reality, introducing any process that hallucinates (read: makes shit up) should not be anywhere near them.
People are free to use LLM plugins to “de-clickbait” headlines as they want, but keep that out of the news process itself.
Just my 2 cents on it.
Slippery slope is a logical fallacy.
speaking even as an “ai” hater, even the worst LLM hallucination isn’t nearly as bad as purpose made clickbait
I tried artifact for a bit, and the blog spam was so bad. Stuff would get promoted that would’ve never made it out of new on Reddit or Lemmy
Yeah, little control over presented articles was what led me to drop it when I was considering reddit alternatives but I’d kill for that de-clickbaiting feature to make way to RSS readers and link aggregators.
The amount of Yanko Design spam in my feed was driving me nuts.
If people use clickbaity titles in the first place it’s cause they work, not because they don’t know how to write informative ones…
Edit: my bad, didn’t know what Artifact was…
It’s for the Artifact users.
Yeah. There are definitely some novel uses and dev efforts. I’m terrified of the day AI data and user interaction data gets monetized though.