The PhD thesis writing process involves dozens of abandoned ideas, code spread across countless repositories, and hundreds of research papers read years ago. AI tools can help synthesize this information. The author shares four specific applications: a research assistant for finding citations, combining code from multiple experiments to improve reproducibility, fact-checking against cited papers, and generating defence questions based on the work. These can be implemented with a basic subscription from one of the AI companies. The applications do not offload mental bandwidth to an AI system and in some cases increased the time spent on the thesis, as they aim to produce better work rather than simply writing it.

The first application, a citation finder, addresses the challenge of recalling specific papers from years of research. The author used a RAG system built with Claude Code, including a BibTeX file with all citations and key papers from the background section. When writing, the author would paste the sentence and relevant papers as a prompt. The author emphasizes not blindly trusting the AI's output and always verifying suggested citations by returning to the original research papers.

The remaining applications covered code consolidation for reproducibility, fact-checking against cited papers, and generating defence questions. All methods were designed to assist rather than replace the researcher's intellectual effort.