You can also just post the 4-5 data items without claiming that this is low or high credibility or bias. Then let the people make the decision. Like this maybe:
“Based on source X, this source media bias is:
Methodology of X is at: “
I find it quite common (and confusing) for certain news types like policy, eg “party A reverses the disapproval to oppose the once-unacceptable ban”
I mean, this article is from 2022, which claims to use seaborn but not really. It really shows their effort, even before the whole AI hype …
https://www.geeksforgeeks.org/how-to-create-a-stacked-bar-plot-in-seaborn/
cuts parts
that’s actually what the underlying method does, as this is extractive summary, hence it mostly cuts and stitches things.
From my naive understanding, this type of method does not use or “understand” context.
The alternative is abstractive summary, which is where LLMs (or even small/medium language models) are good for. But I suspect that would be a controversial choice on lemmy.
I’m also curious. A quick search came up with these. Not sure which one is most reliable/updated
Many things are called “AI models” nowadays (unfortunately due to the hype). I wouldn’t dismiss the tools and methodology yet.
That said, the article (or the researchers) did a disservice to the analysis by not including a link to the report (and code) that outlines the methodology and how the distribution of similarities look. I couldn’t find a link in the article and a quick search didn’t turn up anything.
you should try to ask the same question using xAI / Grok if possible. May also ask ChatGPT about Altman as well
@inspxtr
@lemmy.world