Practical Tips
Our practical experience using AI in professional research
-
Bypassing data protection in LLMs?
Role-playing, poems, academic papers and mathematical encoding: we tested a range of prompt engineering techniques to see whether large language models (LLMs) can reveal information and data that cannot be found using conventional search engines.…
-
Using AI to write better prompts β Part 2
It is not particularly helpful to think of an AI chatbot as an βessentially stupidβ system. This assumption often leads people to write excessively detailed prompts, simply to cover or rule out every conceivable possibility…
-
Using AI to write better prompts
Large language models only produce useful results when instructions and tasks are clearly and precisely defined. The quality of the response directly reflects the quality of the prompt. To develop better prompts, we have often…
-
Analysing large email collections with Copilot in Excel
For a recent project, we needed to analyse more than 120 emails containing around 1,030 individual text items, focusing on the projects and topics they covered. The challenge was to capture information systematically from a…
-
Flagging incorrect assumptions
We tested how clearly different AI tools identify when questions rest on incorrect assumptions or contain errors. To do this, we asked Perplexity, ChatGPT and Googleβs AI Mode three questions each. The first question concerned…