A question related to RAG and "context engineering" from your video "Beyond prompts" on your Substack. I just watched it and wanted to leave a comment but I am not a paid subscriber. I am also very interested in human-AI linguistic interactions and I appreciate your shift in thinking about prompting from info seeking --> info management --> knowledge management -- within the frame of context engineering.
This is already productized in OpenAI's GPTs and Projects, Gemini's Gems, Googlem NotebookLM, and more recently from Antjopic's Skills.
Have you done a comparitive evaluation of these context-based tools?
I was initially very excited about GPTs when they first came out, but I found there was an annoyingly stubborn knowledge drift. Have you put these tools through their paces yet? If so, any thoughts about the context accuracy and consistency of them?
Hi Lance.
A question related to RAG and "context engineering" from your video "Beyond prompts" on your Substack. I just watched it and wanted to leave a comment but I am not a paid subscriber. I am also very interested in human-AI linguistic interactions and I appreciate your shift in thinking about prompting from info seeking --> info management --> knowledge management -- within the frame of context engineering.
This is already productized in OpenAI's GPTs and Projects, Gemini's Gems, Googlem NotebookLM, and more recently from Antjopic's Skills.
Have you done a comparitive evaluation of these context-based tools?
I was initially very excited about GPTs when they first came out, but I found there was an annoyingly stubborn knowledge drift. Have you put these tools through their paces yet? If so, any thoughts about the context accuracy and consistency of them?
Many thanks.