Basic AI Concepts
- 1. Artificial Intelligence (AI)
- 2. Machine Learning (ML)
- 3. Deep Learning
- 4. Generative AI
- 5. Natural Language Processing (NLP)
- 6. Statistical Learning
- 7. Transformers
- 8. Fine Tuning
- 9. Model Validation
- 10. Reinforcement Learning (RL)
- 11. Supervised Learning
- 12. Unsupervised Learning
- 13. System Prompts
- 14. System Roles
- 15. User Prompts
- 16. Zero-shot promptingCurrent
- 17. Multi Shot Prompting
- 18. Templates
Zero-shot prompting
Sep 25, 2025
- In zero-shot prompting the model relies purely on its pretraining to generate a response
- This is the same as a regular user prompt, just using technical terminology.
Example
Write a letter to my landlord explaining why my rent is late.
Zero-Shot vs Multi-Shot
Zero-shot = no examples, One-shot = one example, Few-shot = multiple examples. Multi-shot = lots of examples (10+)
See multi-shot prompting academy post to learn more.
Why it matters
Most everyday GPT use is zero-shot prompting. Understanding this helps you see when adding examples (one-shot or multi-shot) improves results.
Next Article
Continue reading in this category