Introduction to creating video with AI

Summary: On May 28, 2026, Miyako de IT held an online introductory session on AI video generation. Eighteen participants learned Text-to-Video and Image-to-Video basics, how major tools differ, prompt design, and cautions around commercial use.

A beginner-friendly online session

The session welcomed maki, a freelance web designer working in Kyoto, as instructor. It was designed for people who wanted a practical first map of AI video generation rather than a product-only demonstration.

From Text-to-Video to Image-to-Video

Participants reviewed two common approaches: generating a video from text and animating from an image. The discussion covered what each approach makes easier, where results can vary, and how a source image or a detailed prompt changes the output.

Prompt design is a production skill

The group worked through ways to describe a desired scene:

• Subject: who or what appears.
• Movement: what the subject does over time.
• Camera work: zoom, pan, dolly, or a fixed frame.
• Lighting: natural light, neon, cinematic light, or backlight.
• Atmosphere: futuristic, warm, quiet, or dynamic.
• Background: a city, natural setting, room, shop, or office.
• Intended use: social media, advertising, recruiting, or service introduction.

Generative results were treated as drafts to review and improve, not as finished work that needs no editorial judgment.

Commercial use requires checking the current rules

The session also covered the need to review tool terms, subscription conditions, rights in materials, and the requirements of the eventual publication channel. Generated text and marks may not be accurate enough for corporate information, so those elements should be checked or added in editing.

Using reference images and handling consistency

Reference images can make a subject or atmosphere easier to communicate, but results still vary. Participants compared outputs and treated consistency as a review task, not an automatic property of generation.

Commercial use, copyright, and portrait-right cautions

• Check the current commercial-use terms of each tool and subscription plan.
• Confirm rights for images and other source materials entered into a tool.
• Do not assume generated output resolves likeness, copyright, or advertising obligations.
• Review the publication channel’s own rules before distribution.
• Check names, prices, and other factual information during editing.
• Avoid relying on generated text or logos where exactness is required.
• Keep a record of the source materials and production decisions.

The instructor: Kyoto web designer maki

Maki works as a freelance web designer in Kyoto and led the practical introduction for people considering their first AI-video workflow.

About Miyako de IT

Miyako de IT is a Kyoto-based IT community that creates learning and implementation opportunities with practitioners and local partners.

Why AI video generation is a practical skill for engineers, creators, and business development

The skill is not only generating a clip. It is choosing the use case, directing the prompt, checking output, and integrating the result into a truthful communication workflow. Upcoming opportunities are listed on Miyako de IT’s connpass page.

What participants can take forward

The event gave 18 participants a first workflow: choose a small use case, create a prompt or source image, evaluate the output, and edit for the intended audience. For the next practical session, see the short-video PR workshop report.