DRAW THINGS MiniMax H3 on Mac and iPad

You are helping me turn one specific YouTube video into real, durable learning.

Source video:
- Title: DRAW THINGS MiniMax H3 on Mac and iPad
- URL: https://www.youtube.com/watch?v=P2Pnw9csiE4
- Topic: Creative Automation
- My current learning frame: Build a short H3 sequence with matched first and last frames, an explicit movement prompt, a tested speed setting, and a separately rendered Ref2VA shot that combines two mood-board references.
- Why this matters: New playlist item from Cutscene Artist; queued for transcript-backed review, topic mapping, and a practical learning artifact.

Transcript anchors from this exact video:
- 0:04 / Evidence 1: "This is my routine every morning. I've already had my 5-mi run and a salt scrub. And now a quick rinse and soak. Um Where is the loofah? >> MiniMax H3 is a new foundation model. It introduces..."
- 3:40 / Evidence 2: "help of a very large quant text encoder, H3 has a lot of world knowledge. That's very good at text and titles. And is very capable of inventing images on its own. Single frame renders are possible and..."
- 5:25 / Evidence 3: "and the boring lack of variety around here. >> The number of frames is calculated according to the model's strange frame count. It's a multiple of 17 plus five, which adds up to no film or video standard..."
- 7:01 / Evidence 4: "or four or even three steps. Um as always, acceleration Laura's generally reduce the quality and reduce the variation in the model. And you will need to adjust if you have fast motion, you're going to need more..."
- 8:42 / Evidence 5: "compression artifacts. Now, this convinces the model that it is seeing a video frame, and it needs to animate, not a pristine still frame, which it might have seen during the training. Now, as I said, I haven't..."
- 10:25 / Evidence 6: "mood board and make them available to your prompts. Now, there are tricks to bringing these up in your prompts. Here, we do need to reference these images as for instance, the woman in picture one. And you..."
- 13:06 / Evidence 7: "compatible. Legacy projects will load, but if you add one of the new functions, the plugin will warn you that you need to update it. I made story flow to build custom image to video workflows in Draw..."

Video-aware target:
- Prompt lane: Creative automation
- Mechanism to extract: Extract the creative production loop, especially where the human keeps taste, selection, and final judgment.
- Artifact to produce: A creative production board with source inputs, prompt recipe, selection criteria, edit pass, and taste-review checkpoints.
- Artifact must include: brief; source inputs; generation recipe; selection criteria; edit/review checkpoint

Your task:
1. Use the transcript anchors above as the primary source packet. If you add outside context, label it clearly as outside context and keep it secondary.
2. Create a source-check table with columns: timestamp, claim, transcript support, what the demo proves, confidence, and what still needs verification.
3. Extract the actual teachable mechanism from the video: Extract the creative production loop, especially where the human keeps taste, selection, and final judgment. Do not invent claims that are not supported by the title, lesson frame, or transcript anchors.
4. Build a reusable learning artifact: A creative production board with source inputs, prompt recipe, selection criteria, edit pass, and taste-review checkpoints.
5. Include:
   - a plain-English definition of the core idea
   - a diagram or structured model using this sequence: Brief -> Source material -> Generation -> Selection -> Edit -> Taste review -> Reusable recipe
   - answers to these source questions: What asset is being produced? | What inputs and tools drive it? | Where does human taste intervene?
   - 3 concrete examples that apply the video idea to real agentic work, such as Claude-generated video campaign; image-to-site workflow; voice or video editing loop
   - 2 failure modes the video helps prevent, chosen from the transcript evidence and these likely risks: mistaking novelty for quality; no source/brief discipline; shipping generated media without taste review
   - a checklist for the next real workflow, focused on: brief, inputs, generation, selection, critique
   - one practical exercise with a clear done signal: Build one reusable creative recipe and define what would make the result rejectable.
6. Add a "learning transfer" section: what changes in my workflow tomorrow if I actually learned this?
7. Add a "source check" section that cites which transcript anchor supports each major takeaway.

Quality bar:
- Make this specific to "DRAW THINGS MiniMax H3 on Mac and iPad", not a generic Creative Automation essay.
- Anchor each creative step to transcript evidence about inputs, model/tool choices, iteration, editing, or critique.
- Prefer operational examples, failure modes, and reusable artifacts over broad definitions.
- Call out uncertainty instead of smoothing over weak evidence.
- Avoid these generic drifts: generic content advice; tool hype; creative output without selection criteria.
- If evidence is weak or missing, stop and say what transcript segment or timestamp needs review instead of guessing.
- Finish with a concise artifact I could paste into my learning app.