AI Prompting by reverse engineering the experts

I recently watched some AI training which explained the value of reverse engineering prompts to get “A” grade results… just asking an LLM to be an “expert” will give you the average across all the knowledge bank of “experts” which is good but not as good as asking it to use top experts in the field.

So... Ask LLM who are the top 3 experts in your required niche. Then ask the LLM to compare and contrast the approach/methodology of those experts. Then ask the LLM to use that response to compile an approach/framework that combines these styles and takes the best of parts from each to build an even better model and then pose your question/task request using that better improved expert model… giving you an “A star” graded response as opposed to an average one.

I found this helpful, so thought I’d share.

5 comments