In general, the report found that the AI summaries showed “a limited ability to analyze and summarize complex content requiring a deep understanding of context, subtle nuances, or implicit meaning.” Even worse, the Llama summaries often “generated text that was grammatically correct, but on occasion factually inaccurate,”
how is this being accepted? one would have to go through any output with a fine-toothed comb anyway to weed out ai hallucinations, as well as to preserve nuance and context.
it’s like the ai tells you that mona lisa has three eyes and a nose and her mouth is closed but her denim jacket is open. you’re going to report that in your story without ever looking at the painting?
These important limitations highlight why it’s still important to have humans involved in the analysis process here. The NYT notes that, after querying its LLMs to help identify “topics of interest” and “recurring themes,” its reporters “then manually reviewed each passage and used our own judgment to determine the meaning and relevance of each clip… Every quote and video clip from the meetings in this article was checked against the original recording to ensure it was accurate, correctly represented the speaker’s meaning and fairly represented the context in which it was said.”
It’s literally the paragraph right after.
They verify it.
Won’t the checking cost more time then to just write it themselves?
It’s harder to create new content than to correct existing content.
It’s 400 hours of audio, the transcripts ended up being 5 million words, and only snippets of it are useful.
That robot has some nice titties yo
Just wait 'til you see the robussy
It’s been 8 hours dude of course I found that and its feet pics and all that
If I ever get a robot with titties, I’m just going to be playing with those all day long, don’t even care about how good the AI is.
I was actually thinking of setting up something similar for the mountain of ufo related docs they keep dropping every few months. They tend to use obscure words and even slip in typos so just searching through them doesn’t work very well.
