AI video generation starter guide

Summary: AI video generation is easiest to learn by moving from a small prompt experiment to a short, edited communication asset. Miyako de IT’s online sessions introduce Text-to-Video, Image-to-Video, prompt design, and corporate PR use.

Start with two foundations

Text-to-Video creates moving images from a written instruction. Image-to-Video begins with a visual reference and adds motion. Both require a clear idea of subject, setting, composition, camera movement, and intended viewer response.

Our introductory AI-video session report documents a hands-on starting point for these foundations.

Learn prompt design by comparing results

Prompt design is not only about writing more words. Change one condition at a time—such as the product, location, light, action, or mood—and compare the output. This makes it easier to understand what an instruction contributes and what needs to be fixed in a later edit.

Connect generation to real communication

For corporate PR, short generated clips can support product introductions, atmosphere pieces, seasonal messages, and explanations of a service experience. The finished asset still needs an editorial decision: accurate brand information, captions, music, a logo, and a route for the viewer to act.

Our short-video corporate PR workshop report follows this step from generation into publication work.

Major AI-video tools have different strengths

Tools differ in the kinds of generation they emphasize, the length and consistency they can support, their editing controls, pricing, and commercial terms. Choose a tool after defining the communication task rather than treating one product as universally best.

Build prompts around the moving image

Describe the subject, its movement, camera work, lighting, atmosphere, background, and intended use. A useful prompt gives the model enough direction while leaving room to compare several results.

Reference images and consistency are separate problems

A reference image can help with a starting visual, but it does not guarantee consistent faces, products, text, or logos across a generated sequence. Plan an edit and review stage.

Check commercial use, copyright, and portrait rights

Check current service terms, rights in materials entered into the service, likeness concerns, and the distribution channel’s rules before publication.

About the instructor and Miyako de IT

Kyoto freelance web designer maki has led Miyako de IT’s practical introductory session. Miyako de IT is a Kyoto IT community that creates opportunities to learn, connect, and implement.

Continue with practical reports

Read our report on the introductory AI-video session and the short-video corporate PR workshop. Current events are published through Miyako de IT on connpass.

Frequently asked questions

Can I join without experience in AI video generation?

Yes. Some sessions are for beginners and introduce Text-to-Video, Image-to-Video, and prompt design through hands-on practice.

What tools are used?

Tools differ by session. The common themes are generative AI that creates video from text or images, and prompt-design principles that influence the result. See each event report for details.

Can this be used for corporate PR?

Yes. We have held sessions on using short videos for corporate PR, and their reports describe practical publishing and production use cases.

Can I participate online?

Yes. AI video-generation sessions are held online. Each event page states its participation conditions.

How do I join?

Apply through Miyako de IT on connpass. Our guide to joining a mokumoku session explains the basics.