In previous blog post looked at creating tweets and turning a video transcript into a blog post using a large language model such as ChatGPT or gemini: https://llmapplied.blogspot.com/2024/02/using-chatgpt-data-analyst-to-create.html https://llmapplied.blogspot.com/2024/02/google-gemini-video-transcript-to-blog.html In this post I am going to show using ChatGPT to two actions in a prompt :- summarise the transcript but then produce four potential posts with the videos URL and title. In this two prompts will be used. Prompt 1: Will set up general rules for the 'tool': " For the text that will be input. Summarise it to between 250 and 400 words. Also produce four tweets about the text that must include the URL https://www.youtube.com/watch?v=g34jS24JWqw and the title "Swish for Prolog " The second prompt is the actual transcript cut and paste as a prompt. Summary of the text: This video introduces "Swish for Prolog," a tool for experimenting with Prol...
Experiments with various forms of LLMs to improve productivity