When using LLMs to analyze unstructured data like interview transcripts, they often hallucinate compelling but non-existent quotes. To maintain integrity, always include a specific prompt instruction like "use quotes and cite your sources from the transcript for each quote." This forces the AI to ground its analysis in actual data.
Related Insights
After testing a prototype, don't just manually synthesize feedback. Feed recorded user interview transcripts back into the original ChatGPT project. Ask it to summarize problems, validate solutions, and identify gaps. This transforms the AI from a generic tool into an educated partner with deep project context for the next iteration.
To overcome AI's tendency for generic descriptions of archival images, Tim McLear's scripts first extract embedded metadata (location, date). This data is then included in the prompt, acting as a "source of truth" that guides the AI to produce specific, verifiable outputs instead of just guessing based on visual content.
A powerful workflow is to explicitly instruct your AI to act as a collaborative thinking partner—asking questions and organizing thoughts—while strictly forbidding it from creating final artifacts. This separates the crucial thinking phase from the generative phase, leading to better outcomes.
Effective GPT instructions go beyond defining a role and goal. A critical component is the "anti-prompt," which sets hard boundaries and constraints (e.g., "no unproven supplements," "don't push past recovery metrics") to ensure safe and relevant outputs.
Treat AI as a critique partner. After synthesizing research, explain your takeaways and then ask the AI to analyze the same raw data to report on patterns, themes, or conclusions you didn't mention. This is a powerful method for revealing analytical blind spots.
To create a reliable AI persona, use a two-step process. First, use a constrained tool like Google's NotebookLM, which only uses provided source documents, to distill research into a core prompt. Then, use that fact-based prompt in a general-purpose LLM like ChatGPT to build the final interactive persona.
Long conversations degrade LLM performance as attention gets clogged with irrelevant details. An expert workflow is to stop, ask the model to summarize the key points of the discussion, and then start a fresh chat with that summary as the initial prompt. This keeps the context clean and the model on track.
When a prompt yields poor results, use a meta-prompting technique. Feed the failing prompt back to the AI, describe the incorrect output, specify the desired outcome, and explicitly grant it permission to rewrite, add, or delete. The AI will then debug and improve its own instructions.
When an AI model makes the same undesirable output two or three times, treat it as a signal. Create a custom rule or prompt instruction that explicitly codifies the desired behavior. This trains the AI to avoid that specific mistake in the future, improving consistency over time.
To avoid robotic content, use “humanization prompting.” This involves uploading transcripts of your natural speech (from interviews or voice notes) to a custom GPT’s knowledge base, training it to adopt your unique cadence, vocabulary, and style.
