The Art of Prompting
Prompt Strategies
Different problems require different prompting strategies.
There is no single prompt that works for every situation. Effective AI users know how to choose the right prompting strategy based on purpose, context and task complexity.
Lesson Overview
There Is No Single Best Prompting Technique
Ramai pengguna AI mencari satu “magic prompt” yang boleh digunakan untuk semua keadaan. Hakikatnya, prompting lebih menyerupai proses memilih alat yang sesuai untuk tugasan tertentu.
Teknik yang sesuai untuk tugasan mudah belum tentu sesuai untuk tugasan kompleks. Memahami strategi prompting membantu pengguna memilih pendekatan yang lebih tepat, lebih cekap dan lebih bertanggungjawab.
Key Idea
Prompting Is About Choosing the Right Strategy
A good prompt is not only about wording. It is also about strategy. The same task can produce very different outcomes depending on whether you use zero-shot, one-shot, few-shot, role prompting, purpose prompting, Chain-of-Thought or Prompt Chaining.
Prompt Academy 7 Prompting Strategies
Seven Ways to Guide AI More Effectively
01
Zero-shot Prompting
Menggunakan AI tanpa memberi contoh terlebih dahulu. Sesuai untuk tugasan mudah, penerokaan awal atau soalan umum.
02
One-shot Prompting
Memberi satu contoh supaya AI memahami gaya, format atau bentuk jawapan yang dikehendaki.
03
Few-shot Prompting
Memberi beberapa contoh supaya AI dapat mengenal corak, struktur dan standard output dengan lebih konsisten.
04
Role Prompting
Menetapkan peranan AI seperti pensyarah, editor, penyelidik atau jurulatih supaya respons lebih sesuai dengan konteks.
05
Purpose Prompting
Menjelaskan tujuan sebenar sesuatu tugasan supaya AI dapat menghasilkan jawapan yang lebih terarah dan bermakna.
06
Chain-of-Thought
Menggalakkan AI menyusun proses berfikir secara lebih berperingkat untuk membantu tugasan yang memerlukan penaakulan.
Advanced Strategy
07. Prompt Chaining Deserves a Closer Look
Among the seven strategies, Prompt Chaining stands out when dealing with complex tasks that require reasoning, refinement and multiple stages of thinking.
One Prompt Thinking
One Question → One Answer
This may work for simple tasks, but it often fails when the task needs exploration, reasoning, refinement or deeper judgement.
Prompt Chaining
Goal → Prompt → Review → Improve → Final Output
Prompt Chaining breaks a complex task into smaller connected conversations so AI and humans can progressively improve the final outcome.
From Website to Books
Explore the Seven Techniques in Dr Sue’s Books
The seven prompting techniques are introduced in Kuasai Prompt, discussed more systematically in Bijak Kuasai Prompt and explored deeply through Prompt Chaining in Mudahnya Kuasai Prompt.
Kuasai Prompt
Introduces prompting foundations and provides an overview of the seven techniques as part of the learning journey.
Read More →Bijak Kuasai Prompt
Discusses each prompting technique more systematically so readers can understand when and how to use them.
Read More →Mudahnya Kuasai Prompt
A dedicated book on Prompt Chaining and how connected prompts can support complex thinking and better AI collaboration.
Coming Soon →Prompt Academy Insight
Beginners Look for Better Prompts.
Experts build better prompting strategies.
Dr Sue’s Perspective
Choose the Right Strategy at the Right Time
Prompting is not about finding the perfect prompt.
It is about choosing the right strategy at the right time.
Key Takeaways
What You Should Remember
No technique works for everything.
Different tasks require different prompting strategies.
Simple tasks need simple strategies.
Zero-shot or one-shot prompting may be enough for basic tasks.
Complex tasks need connected thinking.
Prompt Chaining is useful when the task requires stages, review and refinement.
Humans still guide the process.
AI can assist, but users decide which strategy fits the goal.
Continue Your Journey
Next Lesson: Prompt in Action
After learning the main prompting strategies, the next step is seeing how these strategies are applied in real-world situations.
Continue to Prompt in Action →