“misrepresent” is a vague term. Actual graph from the study
The main issue is usual… sources. AI is bad at sources without a proper pipeline. They note that Gemini is the worst at 72%.
Note, they’re not testing models with their own pipeline. They’re testing other people’s products. This is more indicative of the product design than the actual models
This graph clearly shows that AI is also shockingly bad at factual accuracy and at telling a news story in such a way that someone who didn’t already know about it to understand the issues and context. I think you’re misrepresenting this graph as being bad about sources, but here’s a better summary of the point you seem to be making:
AI’s summaries don’t match their source data.
So actually, the headline is pretty accurate in calling it misrepresentation.
“misrepresent” is a vague term. Actual graph from the study
The main issue is usual… sources. AI is bad at sources without a proper pipeline. They note that Gemini is the worst at 72%.
Note, they’re not testing models with their own pipeline. They’re testing other people’s products. This is more indicative of the product design than the actual models
This graph clearly shows that AI is also shockingly bad at factual accuracy and at telling a news story in such a way that someone who didn’t already know about it to understand the issues and context. I think you’re misrepresenting this graph as being bad about sources, but here’s a better summary of the point you seem to be making:
AI’s summaries don’t match their source data.
So actually, the headline is pretty accurate in calling it misrepresentation.