Skip to content

The rules of prompting

Prompt any AI. Get exactly what you pictured.

A free, step-by-step prompting course. Learn how to generate the exact image you want, keep characters consistent across video clips, and speak the native language of every major AI platform.

AI prompting workspace with generated image variations

Four disciplines. One vocabulary.

Every AI model responds to structure. Master the underlying rules and they transfer across every platform you touch.

Grid of precise AI-generated image variations

Image precision

Compose prompts like a photographer: subject, lens, lighting, composition, and negative constraints. Stop rolling dice and start directing the frame.

Character continuity

Reference sheets, seed locking, and identity anchors that keep a character recognizable from clip to clip.

Structured reasoning

Role, context, task, format. The four-part frame that turns a vague request into a reliable answer.

Platform fluency

Claude wants context and constraints. Midjourney wants parameters. Sora wants motion cues. Learn the dialect of each engine so your prompts land the first time.

Comparison of prompt styles across AI platforms

The core rules

Seven rules that carry across every model.

01

Be specific, not verbose

Concrete nouns beat adjectives. Say "a red 1967 Mustang at dusk", not "a nice cool car".

02

Give the model a role

"You are a senior copywriter" sets tone, vocabulary, and standards in one line.

03

Show the format you want

Ask for a table, JSON, or five bullets. Models mirror the shape you request.

04

Constrain with negatives

Tell the model what to avoid. "No text, no watermark, no extra fingers."

05

Lock your seed

A fixed seed reproduces a look, so you can iterate on one variable at a time.

06

Iterate in small steps

Change one thing, observe, repeat. Debugging prompts is the whole craft.

A guided path, not a firehose.

Six modules take you from your first prompt to production-grade image and video direction. Each ends with a practical exercise you run yourself.

See the full curriculum
  1. 01

    Anatomy of a prompt

    Role, context, task, format, constraints.

  2. 02

    Text models in depth

    Claude, ChatGPT, and Gemini side by side.

  3. 03

    Generating the exact image

    Midjourney, DALL·E, and Stable Diffusion.

  4. 04

    Character consistency

    Reference sheets and identity anchoring.

  5. 05

    Video & motion prompting

    Sora, Runway, and Kling motion cues.

  6. 06

    Chaining & workflows

    From still to storyboard to finished clip.

New prompting playbooks, monthly.

Get a fresh teardown of a real prompt — image, video, or code — in your inbox. No spam, unsubscribe anytime.