18 ms·
FTA: > This week, there was a viral blog about Gemini 2.0 being used for complex PDF parsing, leading many to the same hypothesis we had nearly a year ago at t
by h0l0cube 2y ago
FTA:
> This week, there was a viral blog about Gemini 2.0 being used for complex PDF parsing, leading many to the same hypothesis we had nearly a year ago at this point. Data ingestion is a multistep pipeline, and maintaining confidence from these nondeterministic outputs over millions of pages is a problem.
- password4321 2y agoYes and per the poster's opening comment: https://news.ycombinator.com/item?id=42966958#42966959 https://news.ycombinator.com/item?id=42966958#42966959
- h0l0cube 2y agoIt seemed you were implying the article was naive to the earlier post, whereas the OP poses itself as a rebuttal. Perhaps a fault of my inference.