Reverse Engineering Responses to Comparative Entity Prompts in LLMs (Chatgpt / Google Gemini)

This pod is a discussion created using Google 's Notebook LM relating to https://www.seolondonsurrey.co.uk's attempt to reverse engineer a response to a comparative entity prompt ("best smartphones") in both Chatgpt and Google Gemini with a simple overarching prompt used in both LLMs. The video looks at the initial prompt (to generate the 'normal' response), comparing and contrasting the 'normal' response and the process outlined used to generate the individual original responses in Chatgpt and Gemini. I've edited some process names out of the discussion.

I cannot be certain the responses discussed do not contain hallucination, but for Chatgpt (the current focus of my research) the results are consistent with my research (on LLMs) hitherto. Subscribe to my newsletter ASAP: https://seoetc.beehiiv.com/subscribe ( I might share the prompts and actual responses)