Close Reading of English-Language Papers is the same task in six steps; the difference before and after comes down to this—not English proficiency, but how many steps I handed over to a graduate-student AI assistant.
Search
Before: I pieced keywords together in a search box, and if a few attempts returned nothing, I assumed no one had done it.
Now: I first write the research question as one sentence and hand it to MuchSci's paper search assistant, which breaks it into search dimensions (Crossref's 180 million records, supports title, author, and DOI search, can filter by year and citation count, and automatically fills in DOI and citation information once a match is found).
Starting out
Before: I opened the abstract and read downward, and when I finished, I still didn't know whether the paper was worth a close read.
Now: I first judge whether the paper is a "background review" or an "original paper strongly related to the topic," then decide how deeply to read it. If you read both types of literature the same way, the result is bound to be: what should be fast can't be fast, and what should be detailed gets glossed over.
Reading itself
Before: I machine-translated the entire text and guessed sentence by sentence, forgetting earlier parts by the time I reached later ones.
Now: Paper translation uses paragraph-level side-by-side reading—the translation and original always line up positionally, with bidirectional Chinese-English academic translation, preserved formulas and terminology consistency, and support for entire PDFs. Use the translation to understand and the original to judge.
Numerical values
Before: I copied numbers from the translation into my notes.
Now: parameters, values, chemical formulas, and conclusions are all checked against the original, including supplementary materials. A translation is not a factual basis; I learned this rule after three days of rework.
Terminology
Before: I looked things up as I read, forgot them once I finished, and the same concept ended up with two translations in my notes.
Now: I record it on the spot in a glossary (Chinese translation + original English term). One more reminder: do not use Chinese translated terms to search databases—you won't find anything, and it's easy to get criticized for it; the original English term and DOI are what you should always keep with you.
After reading
Before: I closed the PDF, the Downloads folder grew messier and messier, and two or three years later I couldn't find anything when I needed it.
Now: the translated draft and key points go together into a personal knowledge base (supports import of PDF, DOCX, PPTX, XLSX, TXT, and MD; personal use by default and not shared). Handwritten notes, whiteboard photos, and files exported from other software go through the personal database layer, which has a broader range of sources; reports generated by AI note-taking can also be stored there.
In one sentence: before, I put effort into "understanding"; now, I put effort into "what's left after reading."