Excellent post. I'm so happy I've found your Substack. I'm PARTICULARLY happy to see your posts are written in a supportive way to writers wanting pragmatic knowledge on using AI tools in their writing process. Versus - constantly running down the use of AI tools. People don't seem to realize that even Grammarly uses a very complex AI backend to the writing tools they provide. So thanks. I'm finding lots of good tidbits from you. I'm making my way through many of your posts today with a smile on my face and a notepad nearby.
I was just about to respond with the comment that Daniel Nest makes below. It does help to ask the LLM for 'more questions' or to go deeper with questions as follow up - or series of followup prompts to force it dig deeper and uncover something new for me.
Sometimes, after giving the initial batch of answers, I add extra loops by appending something like "If you have more questions based on my answers, please ask them."
This forces chatbots out of what they've been trained to do---spitting out an answer---and into a more back-and-forth mindset that benefits both parties.
Excellent post. I'm so happy I've found your Substack. I'm PARTICULARLY happy to see your posts are written in a supportive way to writers wanting pragmatic knowledge on using AI tools in their writing process. Versus - constantly running down the use of AI tools. People don't seem to realize that even Grammarly uses a very complex AI backend to the writing tools they provide. So thanks. I'm finding lots of good tidbits from you. I'm making my way through many of your posts today with a smile on my face and a notepad nearby.
I was just about to respond with the comment that Daniel Nest makes below. It does help to ask the LLM for 'more questions' or to go deeper with questions as follow up - or series of followup prompts to force it dig deeper and uncover something new for me.
Keep the good stuff coming!
Tom
AI for lifelong Learners
Hear hear!
This is exactly the approach I've shared in this older post of mine: https://www.whytryai.com/p/two-methods-ai-chatbots-prompt-themselves
Sometimes, after giving the initial batch of answers, I add extra loops by appending something like "If you have more questions based on my answers, please ask them."
This forces chatbots out of what they've been trained to do---spitting out an answer---and into a more back-and-forth mindset that benefits both parties.