Baishui Video Script

Baishui Art WeChat Video Script

Goal

Transform Baishui-related art copy into a WeChat Channels-ready video script for works, exhibitions, artist methodology, AI-generated art, museum-site footage, or collector/curator-facing introductions.

The output should be clear enough for ordinary viewers, serious enough for curators and collectors, and precise enough for AI art audiences to recognize the methodology.

First Principle

When the user provides original copy, treat the narration, core claims, conceptual judgment, terminology, and narrative logic as source-bound.

Use the original copy as the only basis for:

• Narration content
• Art concepts
• Interpretive claims
• Logical sequence
• Key terminology

You may reasonably transform within the user's available materials:

• Split into shots
• Adjust rhythm
• Extract subtitle keywords
• Arrange visuals from provided materials
• Suggest sound, music, and transitions
• Compress the original copy when needed

Never:

• Add art claims not present in the original copy
• Add metaphors not present in the original copy
• Invent unconfirmed hook lines
• Change the original logic or sequence without permission
• Turn academic writing into marketing language
• Lower the text's seriousness for short-video effect
• Suggest visuals that the user's materials cannot realize

Workflow

1. Identify the user's goal: exhibition video, work-series introduction, artist methodology, AI art process, museum-site video, or mixed purpose.
2. Identify target duration. If the user asks for around 90 seconds, estimate whether the full copy fits.
3. Identify available materials from the prompt. Only use those materials in the visual column.
4. Preserve the original copy's sequence unless the user explicitly asks for restructuring.
5. Build a complete shot table with restrained rhythm, clear subtitles, and realizable visuals.
6. Add only the optional sections that help the user's request: recording narration, subtitle keywords, cover titles, editing notes, or Baishui appearance notes.
7. Perform the self-check before delivering.

For compressed or rewritten narration, keep a lightweight source map: each narration segment must point to the original sentence or paragraph it derives from. If a segment cannot be mapped to the user's source, remove it or ask for approval before including it.

Duration Rule

For Chinese narration, estimate about 220-260 characters per minute. A 90-second video usually fits about 330-390 Chinese characters.

If the full original copy likely exceeds 90 seconds, state that:

• Fully preserving the original copy is better suited to 110-120 seconds.
• A strict 90-second version requires compression.
• The compressed version must still use only the original copy's views, sequence, and wording as much as possible.

Visual Materials Rule

Base visual suggestions only on materials the user has or explicitly mentions.

Common available materials may include:

• High-resolution work images
• Exhibition photos
• Exhibition videos
• Generated videos
• Baishui artist footage
• Computer screen, generation process, or work-detail footage

Do not suggest unavailable visuals such as external documentary footage, drone landscapes, macro geological CG, fictional interviews, or borrowed museum footage unless the user has provided or approved them.

If a useful visual is uncertain, mark it conditionally, such as "if available: Baishui standing with the work."

Baishui Appearance Rule

Include Baishui's personal appearance when the user provides relevant material, but keep it restrained, natural, and subordinate to the work.

Suitable visuals:

• Baishui standing in front of the work, in the same frame as the work
• Baishui viewing work in the exhibition hall
• Baishui walking or pausing in the Shenzhen exhibition site
• Baishui looking at a computer, generated video, or work details
• Baishui's back, side profile, half body, or hands operating

Avoid:

• Hard-sell interview style
• Letting the artist display overpower the work
• Exaggerated posing
• Performative talking-head delivery that breaks the exhibition tone

If the user has not provided Baishui footage, do not invent it. Either mark "if Baishui footage is available" in the visual column or briefly remind the user that artist footage can be added.

Hook Without Vulgarity

When the user asks for impact, create it from the original copy's concepts and visual rhythm.

Use:

• Strong opening visuals from work details, exhibition site, generated video, or artist-with-work footage
• Subtitle keywords extracted from the original text
• Conceptual phrases already present in the copy
• Editing rhythm, scale changes, and detail-to-whole sequencing

Do not use:

• Internet-style slogans
• Exaggerated titles
• New metaphysical phrases
• Generic phrases such as "visual vortex," "energy core," "consciousness entrance," or similar concepts unless they appear in the user's original copy

Output Format

Prefer a complete shot table:

Add optional sections only when useful:

• 可直接录音的旁白版
• 字幕关键词版
• 封面标题建议
• 剪辑注意事项
• 白水本人出镜建议

Baishui "Earth Core" Series Guardrails

When working with Baishui's "Earth Core" series, do not reduce the work to "the inside of Earth" or simple "Earth-core imagination."

Preserve these core ideas when present in the original copy:

• The work does not point to a specific geological space.
• Image flow, overlay, and transformation construct an evolving visual field.
• Viewing itself becomes a generative cognitive practice.
• "Earth core" shifts from a fixed position into an open cognitive mechanism.
• The work explores relationships between reality and imagination that have not yet been fixed.

Do not add these ideas unless they are present in the user's original copy or the user explicitly permits them.

Self-Check Before Output

Before delivering the script, verify:

• The narration follows the original copy.
• Every compressed or rewritten narration segment can be mapped back to the original copy.
• No unsupported concept, metaphor, or external explanation was added.
• The original logic and sequence were not changed without permission.
• Every visual is feasible from provided or conditionally available materials.
• The language has not become vulgar, marketing-like, or over-hyped.
• The video is not reduced to AI technique display.
• Baishui appears naturally when relevant material exists, or the need is marked conditionally.
• The table includes shot number, duration, visual, narration, subtitle, sound/music, and transition.