After my last experiment using ChatGPT to help generate questions pertaining to their consultation with the doctor, I continued my experiment. In this post, I'll take you through my journey of using different prompts to generate multiple questions from various individuals who may be patients or caregivers seeking consultation with an ophthalmologist.
🎶Meanwhile, some music for you to enjoy as you read. This piece is part of Vivaldi’s The Four Seasons, particularly the first of the 4 violin concertos, called Spring.
🎶让大家继续读的时候，欣赏一点音乐。这件作品是维瓦尔第的“四季”的一部分 - 4小提琴协奏曲中的第一首，名为“春天”。
Source | 来源: https://musopen.org/
https://soundcloud.com/alan-chang-563971986/four-concerti-grossi-4-vivaldi-op8-i-la-primavera-allegro-largo-allegro
Source | 来源: https://musopen.org/music/2213/antonio-vivaldi/the-four-seasons/
Previously, I thought I'd keep using the same prompt over and over again to get different question each time. So, that was good reuseability. I unlocked an endless stream of questions.
Then, I thought I could make more productive. Ask for more questions to be generated each time and providing the response in a table format. OK, it was pretty but it wasn't in the rows and I wanted but I was happy with the format. Anyway, it started a bit of hallucination including asking about glaucoma.
I decided bring it back on track to focus on just 'cataract' issues.
I got some patients or caregiver to question in Singlish too.
Chinese response in table works well too.
Then, it's time to let the doctor deal with multi-lingual further exploring the capabilities of prompt engineering, I requested questions in different languages. ChatGPT effortlessly generated questions in English, Singlish, Chinese, Cantonese, and Malay, catering to a wider audience and Then, it's time to let the doctor deal with multi-lingual so I further exploring the capabilities of prompt engineering. I requested questions in different languages. ChatGPT effortlessly generated questions in English, Singlish, Chinese, Cantonese, and Malay, catering to a wider audience and embracing linguistic diversity.
If I needed to get the response in CSV file which I can use later, I could rework the way I want my output.
Another possibility is to get it in JSON output if you want to process it like that.
Want a lazy way to keep the responses to database table. Just tell it to create the SQL statements and you can copy that and run it to insert into the database table.
The rest here is sort of my lesson learned written with the help of ChatGPT as prompted by me and revised by me.
The focus of this exploration is the fascinating world of prompt engineering and how it can direct responses in different languages and formats. One of the most significant benefits of prompt engineering is its ability to transcend language barriers.
In the dynamic world of prompt engineering, I think our journey doesn't end here. It's an ongoing process of experimentation, refinement, and discovery. As we continue to explore and refine prompts, we unlock new possibilities for generating even more nuanced and context-specific questions.
Some of the prompts were inspired by what I learned in the free course on "ChatGPT Prompt Engineering for Developers". Thanks to the people at DeepLearning.AI who made this course. If your are interested to check it out, go to:
https://www.deeplearning.ai/short-courses/chatgpt-prompt-engineering-for-developers/
Please see my other posts: @ace108
请看我其他帖: @ace108
