Top news
Investing in the public benefit OfficialAnthropic NewsOct 18Popular Musk Pledges SpaceX Shares to ‘Every Kid in America’ WebBloomberg TechnologyOct 19Popular Websites are closing their doors to AI agents WebThe Verge AIOct 17 This is the new ChatGPT Voice, powered by GPT Live OfficialOpenAI on YouTubeOct 18 How AI giants will reshape global philanthropy WebAxios AIOct 19 Gemini 4 Argon Really Surprised Me WebTwo Minute PapersOct 19 Google Gemini 4 agents just beat Minecraft with no human intervention WebEngadget AIOct 19 Who could it be? 🧐 OfficialGoogle on YouTubeOct 18 Anthropic gives outside AI watchdog “employee-like” access WebArs Technica AIOct 18 What are self-replicating ‘AI worms’, and why is Anthropic worried about them? WebThe Guardian AIOct 18
GlossaryModelsIntermediate

Prompt Engineering

Also written: Prompting, Prompt design

Designing and refining input prompts to guide AI models toward desired outputs.

In plain terms

Prompt engineering is about learning how to talk to AI models. It means giving them clear, specific instructions so they can give you the best possible answers or create the content you want.

Think of it like asking a very smart assistant for help. If you ask a vague question, you might get a vague answer. But if you are precise and give details, you'll get a much better result. For example, instead of asking an AI to 'write about dogs,' you could ask it to 'write a 500-word blog post about the benefits of adopting senior dogs, aimed at first-time dog owners.'

The expert version

Prompt engineering involves the iterative design, refinement, and optimization of input prompts to steer generative AI models toward producing specific, accurate, and relevant outputs. This process requires understanding the model's capabilities and limitations, and often involves techniques like few-shot learning, chain-of-thought prompting, and careful instruction formatting.

The effectiveness of prompt engineering can significantly impact model performance, influencing factors such as response quality, factual accuracy, and adherence to desired stylistic constraints. Tradeoffs exist between prompt complexity and the computational resources required for inference, as well as the potential for prompt injection vulnerabilities.

Why it matters now

Prompt engineering is a key focus as companies like OpenAI and Anthropic release best practices and courses. The goal is to help users and businesses get more accurate and useful responses from AI models. This refinement process is crucial for practical applications, from marketing to business operations.

Where it shows up 12 items, newest first
How each company uses it
Anthropic Claude
Anthropic offers GitHub courses on prompt engineering and shares tips for improving Claude's business performance.
OpenAI ChatGPT
OpenAI provides best practices for prompt engineering with ChatGPT and hosts workshops on its practical use.
Related terms

Generative AILarge Language ModelsAI AlignmentFew-shot LearningChain-of-Thought PromptingAI Safety

Often confused with
Fine-tuning
Fine-tuning retrains a model on new data, while prompt engineering
AI Training
AI training involves teaching a model from scratch or with large da
Natural Language Processing
NLP is a broader field of AI focused on language un
Origin

The term emerged from the AI research community as large language models became more capable and sensitive to input phrasing. Early work focused on understanding how different prompt structures affected model outputs.

Try this about 5 minutes

Open ChatGPT or Claude and ask it to write a short story about a robot who discovers music. Then, try again, but this time ask it to write the same story but from the perspective of a cat who is annoyed by the robot's music. Notice how the specific instructions change the output.

Try it: Playbook entries that use thisPlaybook →

The Playbook: reusable how-tos, prompts, configs and tips pulled out of the news.

Questions people ask

What makes a prompt 'good'?
A good prompt is clear, specific, and provides enough context for the AI to understand your request and generate a relevant response.
Do I need to be a programmer to do prompt engineering?
While advanced prompt engineering can involve technical skills, basic prompt engineering is accessible to anyone who can write clear instructions.

First seen on Subvolts Dec 14, 2021. Updated Oct 19, 2026. Written by an AI model from the items above and reviewed against them weekly; report an error.