how i make AI videos that print me $12k per day
Published
Should I watch this?
Skip to the real contentYour full summary is ready below. Read on while we finish your watch recommendation.
Your summary is ready. The watch recommendation is next.
Worth watchingSkim
The reference-image-to-Claude/MCP walkthrough and missing-variant fix offer practical ideas, but repeated realism advice, Instagram-gated skills, and unverified earnings pitches make selective viewing more worthwhile.
The main answer
The video’s practical answer is: make realistic AI product videos by combining strong reference images, a curiosity-driven opening, and specific generation prompts—not just generating random clips. The demonstrated workflow is: 1. Collect clear product photos, preferably from several angles. 2. Research competitors on Instagram. Find an advantage: better visuals, a fresh concept, or a different product variation. 3. Download a reference video using SnapInsta and have Claude analyze its frames using the creator’s supplied skills. 4. Supply an avatar image, background image, and product photos. Generate a convincing starting frame. 5. Connect Claude to the platform called “Xfield” in the transcript through MCP. Describe the dialogue, actions, lighting, reactions, and camera movements; generate roughly 10-second, 1080p clips. 6. Correct missing elements, assemble the clips, and use readable, well-positioned text. This demonstrates a production method, not a proven formula for virality or income.
Key takeaways
- Research before producing: the creator recommends avoiding products where you cannot identify a meaningful advantage over existing content.
- The opening should create curiosity or interrupt an expected pattern. Attractive footage alone does not guarantee that viewers keep watching.
- Realistic starting images, lighting, reactions, backgrounds, and clear prompts are the tutorial’s central production priorities.
- Iteration is still necessary: the demonstration initially omits requested product variations, which the creator generates afterward.
- The claimed $3,000 per million views refers to product-sales performance, not a platform payout. Applying that rate to 14.6 million views gives $43,800, but does not prove the featured video earned that amount.
- The example produced during the tutorial is not shown being published or generating sales. Claims that it would perform well remain predictions.
- Using Pinterest photos of real people and downloading competitors’ content raises licensing, likeness-consent, and originality concerns. The transcript does not address these or disclosure of synthetic product demonstrations.
Summary
The creator demonstrates how to research and generate realistic AI product-promotion videos, emphasizing strong hooks and visual quality. Useful production advice is mixed with unverified earnings claims and a closing pitch for mentorship.
What’s being promoted?
The clearest commercial promotion is the creator’s mentorship, linked in the description. The pitch promises help choosing a “winning product,” developing its strategy, and producing content at scale. Pricing, terms, and independently verified outcomes are not supplied. “Xfield” receives repeated favorable exposure as the generation platform. Claude is recommended for brainstorming, analysis, and connected generation; SnapInsta, Pinterest, and CapCut/TikTok fonts are supporting workflow recommendations. The transcript does not establish sponsorship, affiliate payments, or ownership relationships with these tools. There is strong evidence of a customer-acquisition agenda alongside the tutorial: viewers must message “prompt” through Instagram Stories to obtain useful skills, are directed to another video, and then receive the mentorship offer. This builds contact with potential customers while positioning the creator as the person who can close their alleged “knowledge gap.” The persuasion relies on spectacular student earnings, claims that beginners succeed quickly, a claimed million-dollar client month at 70% profit, and Q4 urgency. These examples may be genuine, but the transcript cannot verify them or establish how representative they are. It does not adequately distinguish sales from take-home profit, explain costs and failures, or support promises of repeatedly going viral. The commercial motive is explicit; deliberate deception or undisclosed tool sponsorship remains unproven.
