Guide · Researching

How to use AI for research

AI genuinely accelerates research at four points: finding literature, pressure-testing a method, exploring data, and drafting prose. It cannot be the authority on any factual claim, and it fabricates citations in a form that looks correct. Used with a verification step attached, it is the largest practical gain research workflows have had in years.

Last reviewed Who this is for: Doctoral students, academics and analysts who want the speed without becoming the person who cited a paper that does not exist.

How long does it take to learn?

A week to change how you read literature. A term to trust it near your own methodology, which is the right pace: the failure mode here is not slowness, it is a retraction.

The fastest path, in order

  1. 01

    Use it to map a field, not to state facts

    Ask what the debates are, who disagrees with whom, which papers everyone cites. It is very good at shape and unreliable on specifics, and directing it at the thing it is good at is most of the method.

  2. 02

    Find papers with tools built to find papers

    Elicit, Consensus and Semantic Scholar search real indexed literature. A general chatbot asked for references is guessing what a reference would look like, and it will produce one.

  3. 03

    Verify every citation, mechanically

    DOI, title, authors, year, and the sentence you are citing it for. Fabricated references are correctly formatted and plausibly titled, so eyeballing does not catch them. Only lookup does.

  4. 04

    Use it against your method, not for it

    Ask what a reviewer would attack, what confound you have missed, what the counter-hypothesis is. As an adversary it is excellent. As the author of your design it is a liability.

  5. 05

    Let it write code, not conclusions

    Analysis scripts, cleaning, plots: strong, and checkable by running them. Interpretation of what the numbers mean stays yours, and that boundary is the one worth holding hardest.

  6. 06

    Draft with it, and read every line

    It is good at structure and transitions and prone to confident filler. Know your institution's disclosure rules and your journal's policy before submission, not during.

Tools worth your time

ToolWhat it is for
ElicitLiterature search and extraction over real papers, built for systematic review workflows.
ConsensusSearches findings across papers and shows where the evidence agrees and where it does not.
Semantic ScholarFree index with citation graphs, the fastest way to confirm a paper exists.
Connected PapersVisual map of what a paper is related to, useful for finding what your search missed.
NotebookLMGrounds answers in documents you upload, so it quotes your sources instead of its memory.
ZoteroReference manager. Unglamorous, and the thing that makes citation checking a lookup.

Grounded tools that read documents you provide are categorically safer than open-ended chat for anything factual, because the answer is tied to a text you can open. That single choice removes most of the fabrication risk before any checking begins.

Mistakes that cost people weeks

The track that teaches it

Why this is the fastest way to actually get there

The academic conversation about AI is split between people who ban it and people who use it uncritically, and neither produces a method. What is missing is where exactly it helps, where it must not be trusted, and what checking looks like as a procedure.

  • Organised by research stage, so it maps onto the work rather than onto a tool list.
  • The verification procedure is explicit and short enough to run on every citation, which is the only version anyone actually runs.
  • Written for people whose reputation depends on being right, which sets the standard for everything the track recommends.

Questions people ask

Partly, and it depends on your institution's rules, which now exist and differ. Drafting, structuring and editing are widely accepted with disclosure. Generating findings or citations is not, anywhere.

Because it generates plausible text rather than retrieving records. A citation has a strong, learnable shape, so producing a convincing fake is exactly what the process does well. Tools that search real indexes avoid this.

Elicit or Consensus for searching real papers, Semantic Scholar for confirming they exist, NotebookLM for questioning documents you already have. A general chatbot is the wrong tool for this specific job.

Detectors are unreliable in both directions and should not be your concern. Disclosure policies are, and they are increasingly explicit. Follow your journal's and your institution's rules and the question stops mattering.

Literature review, methodology, analysis and academic writing with AI, organised by research stage, with the citation verification procedure taught as a step rather than a warning.

Vincoria opens soon.

One email when the doors open. Nothing else.

Founding members hear first. No spam.

Used only to tell you when Vincoria opens. One click to leave, from any email. Privacy policy

Next stepStart Lesson 01 · freeDevelop with AI · 9 min