After surveying the landscape of AI and productivity tools for researchers, here’s a practical toolkit recommendation for PhD students and early-career researchers. This is opinionated and specific — not a comprehensive list of every available tool.
The Core Stack (Everyone Should Use These)
Zotero (free): Reference management. Non-negotiable. If you’re not using Zotero, start today. Browser extension, Word/LibreOffice integration, shared group libraries, free sync for 300 MB.
Claude or ChatGPT (~$20/month): General AI assistant for writing feedback, argument structuring, document translation, code explanation. Claude is preferred for long documents; either works for most tasks.
Semantic Scholar (free): Academic search with AI features (TLDR summaries, citation context, author tracking). Better than Google Scholar for research-specific needs.
Obsidian (free for local): Personal knowledge base for linking reading notes. Takes 2–3 months to see the value; after 12 months it becomes indispensable.
Discovery Layer (Add When You Start a Literature Review)
ResearchRabbit (free): Citation network visualization. Use once per new topic to map the field. Connected Papers (5 free graphs/month): Bibliographic coupling visualization — complementary to ResearchRabbit. NotebookLM (free): Upload your PDF collection and query across it. Especially useful for synthesis before writing chapters.
Specialized Tools (Add When You Have a Specific Need)
Elicit ($12/month): Systematic literature extraction. Use for one systematic review, then evaluate whether the subscription continues to justify itself. SciSpace (free tier available): Reading tool for dense papers. Use the browser extension on papers you find in the wild. Cursor ($20/month): AI code editor. Essential if you write Python/R regularly.
What to Skip
Tools you don’t need: AI journal recommendation services (your supervisor knows which journals are appropriate), AI thesis generators (the ideas must be yours), subscription services that duplicate free tools you already have.
The Single Most Important Habit
Whatever tools you use: write the hard parts yourself. AI makes the mechanical work faster; doing the hard intellectual work yourself is what builds the expertise that matters for your career. Tools compound over time only if the expertise underneath them is growing too.
