A Chinese academic post that went viral — “I’ve decided to let Claude accompany me through writing my 200,000-character PhD dissertation” — captured something real: AI is changing the dissertation writing experience, not by writing it, but by changing what writing a dissertation feels like.
The Isolation Problem AI Addresses
Dissertation writing is isolating. You spend months writing for an audience that won’t read until much later, getting feedback on complete chapter drafts every few months. AI changes this: you have a reader for every draft at every stage, with consistent availability and no social cost to asking basic questions.
Chapter Outline Development
Before writing a chapter, describe the chapter’s purpose, its place in the dissertation narrative, and the key claims it needs to make. Ask Claude to draft a detailed outline. This outline forces you to think through the argument before writing prose. Revise the outline until it accurately reflects what the chapter needs to accomplish. Writing from a worked-out outline is faster and produces cleaner prose than writing exploratorily.
The Daily Writing Ritual
Some PhD students have developed a daily startup ritual: paste yesterday’s writing, ask “Where did I leave the argument? What should the next section accomplish?” This provides continuity across writing sessions — re-entering a complex argument after sleeping is cognitively expensive. A brief AI summary of where you left off speeds this re-entry.
Managing Advisor Feedback
After receiving supervisor feedback, paste the feedback comments and the relevant draft sections. Ask Claude: “My supervisor said X about this section. What specific changes would address this concern? What does the supervisor seem to want that I might be missing?” This converts vague critical feedback into concrete revision tasks — particularly valuable for PhD students whose supervisors communicate in implicit academic norms that aren’t always spelled out.
The Authorship Line
Every word in your dissertation represents your scholarly contribution. AI helps you produce your contribution more efficiently — it doesn’t contribute to the scholarship itself. The ideas, the research, the interpretation of data, the scholarly synthesis: these are yours. Be transparent with your institution about how you’re using AI tools; policies are evolving rapidly and vary by university and country.
