Summary of "'프롬프트 무료로 드립니다' 배달일 그만두고 AI경제 유튜브 채널 만들어 월 3000씩 버는 남자"

High-level summary (business focus)

Case: An individual identifying as “Vinipa” transitioned from high-effort manual work (motorcycle delivery and part-time jobs) to operating multiple AI-driven YouTube channels. These channels focus on animated, educational videos about economics and health. He developed a repeatable, automated content-production system that scaled to substantial ad-revenue profits and multiple income streams across Korea, Japan, and Taiwan.
Core strategy (condensed):

Scout fast-moving trends in the U.S., localize and adapt them to target markets (Korea → Japan/Taiwan), produce animated educational videos using AI, iterate thumbnails/formats from top-performing channels, and scale by running multiple channels with automated production.

The process is an explicit 3-step workflow combined with automation.

Trend discovery & topic sourcing
- Monitor trending channels and topics overseas (especially the U.S.) to identify high-performing formats.
- Quickly localize successful topics to target markets (Korea first, then Japan/Taiwan).
Content generation (three-step creation flow)
- Script generation
  - Use Google Opal plus custom prompts to collect the latest news articles and auto-generate a structured script.
- Scene generation & audio
  - Feed the script to a custom video-editing program that segments it into scenes (up to ~50).
  - Auto-generate images per scene via AI and produce matching AI voiceover (function: “Generate Full Audio”).
- Thumbnail & packaging
  - Use an AI tool (Zenspark) to mimic top-performing thumbnail styles.
  - Capture top thumbnails as references and prompt the tool to produce localized animated thumbnails with consistent characters/branding.
Publish & iterate
- Upload videos and monitor performance metrics; use low-subscriber / high-view signals to identify “hot” topics early.
- Run a portfolio of channels so winners can be scaled while other channels serve as experiments.

Built a macro program that automates the end-to-end workflow: copies scripts from source videos (with timestamps), inserts them into the internal pipeline, generates images, audio, and video, and exports finished videos ready for upload.
Treats AI systems as “employees”; his role shifts to management and oversight instead of hands-on production.
Channel portfolio approach: operates 5 main channels plus ~10 test channels; aggregate output up to ~10 videos per day.
Uses consistent characters/branding (example: Shiba Inu / “Zolaman” style) for recognition across markets.
Claims productivity improvements: per-video hands-on time reduced from ~5–6 hours to ~1 hour, and further toward near-zero with macros.

Risk: Channels that directly copy news articles or show low originality can be sanctioned (earnings suspended or channel deletion).
Mitigation tactics:
- Use animation and educational framing; add creative/original elements.
- Avoid straight copy/paste of news items; localize content and add commentary or analysis.
- Leverage YouTube’s current favorable distribution of animation/educational formats.

Google Opal — script generation and news aggregation
Custom program / internal video editor — script segmentation into scenes and batch image/audio generation
Zenspark — AI thumbnail generation and “super agent” prompt templates
Macro automation (custom) — orchestrates the entire pipeline and simulates user actions end-to-end
Various AI image and voice generators (unnamed models) for visuals and narration

Peak monthly revenue: ~50,000,000 KRW in the best month.
Typical/average monthly income: 30–40 million KRW.
Floor monthly profit since pivot: never below 10 million KRW.
Example per-channel and daily figures:
- One Japanese channel: recorded daily profits of 1,000,000 KRW (on the 28th) and 1,700,000 KRW (on Jan 7).
- Another channel: 31 million KRW in August and 20 million KRW in September; often around 10 million KRW monthly.
Aggregate examples:
- “With me and two other channels, it’s almost 10 million” (implying aggregated channel revenue exceeding 30 million KRW across multiple channels in context).
Output cadence:
- Some channels publish about 1 video every 1–3 days.
- Total production capacity across the portfolio: ~10 videos per day.
User outcomes / case examples:
- Sister-in-law: 10–20 million KRW profit after using the system.
- Grandfather (born 1959): two channels earning 8.2M KRW and 5.2M KRW.
- A beginner: gained 1,000 subscribers and started earning within ~3 weeks.
- Multi-channel example: three revenue channels reached 5M, 5M, and 3M KRW (total 13M) in ~3 months.

Scout U.S. trending formats and quickly translate/adapt them for domestic audiences, then expand to neighboring markets.
Favor animated, educational content (economy and health) because YouTube currently promotes these formats.
Use a script generator (Google Opal or similar) to:
- Aggregate news, plan topics, and automatically write scripts.
- Segment scripts into scenes for automated image and voice generation.
Apply thumbnail-mirroring:
- Capture thumbnails from top-performing videos and feed them to Zenspark with prompts to produce similar high-conversion thumbnails.
- Swap characters and localize branding for different markets.
Avoid direct copying: add creativity, educational value, and localization to reduce policy risk.
Build a portfolio of channels: run multiple main channels and experiments; scale winners and retire or pivot underperformers.
Automate repetitive tasks via macros and pipeline orchestration to scale productivity and reduce hands-on time.
Mentor and teach others; distribute tooling (a free program was mentioned) to accelerate adoption and create referral/success stories.

Productize the production pipeline (internal tooling + macros) to make content creation repeatable and scalable.
Run parallel experiments: test multiple channels and formats, measure views/subscribers/profitability, then clone success across languages/markets.
Time leverage: use AI to multiply effective output hours and reduce labor intensity compared to manual work.
Management posture: act as a curator/manager of AI processes rather than a content worker. Invest time in systematization to gain personal freedom (example: more time with family).

YouTube enforcement risk if content is unoriginal or copies news sources — possible sanctions include earnings holds or channel deletion.
Must ensure educational/creative packaging to stay within YouTube policy and benefit from algorithmic support.
Quality control is essential: validate AI outputs and manage brand consistency and factual accuracy, especially for news and economic content.

Rapid ramp: claim of going from delivery work to up to 50M KRW/month within about one year.
Short-term results for learners: some beginners gained subscribers and revenue within ~3 weeks to 3 months using the system/tools.

Vinipa — subject, AI YouTuber, former motorcycle delivery worker (referred to as 대표님 / CEO in the interview).
PD — interviewer/producer who filmed the session.