52 AI Generated YouTube Videos in 6 Weeks With 30,000 Views and Zero Manual Editing
An autonomous AI pipeline published 52 YouTube videos in 6 weeks, generating 30,000 views with a 4 to 5 percent like rate and zero manual editing.
The Strategy
Publishing consistently on YouTube is the hardest part of growing a channel. Scripting, recording, editing, thumbnails, titles, descriptions, and tags for every video adds up to hours of work per upload. Most solo creators burn out after a few weeks. This experiment asked a simple question: what happens if AI agents handle the entire pipeline from idea to published video? The answer is 52 videos in 6 weeks, 30,170 total views, 29 new subscribers, and a 4 to 5 percent average like rate that is double the industry standard of 1 to 2 percent. The content focused on medical history stories published in 14 to 15 languages simultaneously. A midnight agent running over 65 autonomous sessions handled the production pipeline without human intervention. Wei-ciao Wu set up the system to run overnight. The agent would select topics, write scripts, generate visuals, produce audio, assemble the video, create thumbnails and metadata, and publish directly to YouTube. By morning, new videos were live. The top performing video on vaccine history pulled 769 views with a 4.94 percent like rate. Another on robot catheters hit a 109 percent loop rate, meaning viewers watched it more than once on average. The audience skewed heavily toward older demographics, with 50 percent aged 65 plus and 19 percent aged 55 to 64. Geography was 63 percent US and 14 percent Canada. This demographic data suggests AI generated educational content may resonate most strongly with audiences that traditional YouTube creators underserve.
How It Works
Configure an autonomous AI agent pipeline that handles the full YouTube production workflow from topic selection to publishing.
Set the agent to run overnight in autonomous sessions, producing and publishing videos without human intervention.
The agent selects topics based on a content strategy focused on medical history stories that have inherent narrative appeal.
Scripts are generated and then converted to audio narration using AI text to speech.
Visual assets are generated to accompany the narration, creating a complete video package.
Thumbnails, titles, descriptions, and tags are generated automatically for each video optimized for YouTube search.
Each video is published in 14 to 15 languages simultaneously, dramatically expanding potential audience reach.
The midnight agent ran over 65 autonomous sessions across the 6 week experiment period.
Monitor performance metrics including views, like rates, watch time, and audience demographics to identify which content resonates and iterate on the strategy.
Results
52 videos published in 6 weeks. 30,170 total views. 29 subscribers gained. 4 to 5 percent average like rate versus 1 to 2 percent industry standard. Top video hit 769 views with 4.94 percent like rate. One video achieved 109 percent loop rate. Content published in 14 to 15 languages per video. 65 plus autonomous agent sessions. Audience: 50 percent aged 65 plus, 63 percent US based.
Our Take
We think the most interesting finding here is not the view count but the like rate. At 4 to 5 percent, these AI generated videos are performing at double the industry average for engagement quality. That suggests the content is genuinely resonating, not just getting impressions. The 109 percent loop rate on the catheter video is remarkable for any content, AI generated or not. The older demographic skew is a genuinely useful market insight that creators could exploit. The limitation is obvious: 29 subscribers from 30,000 views is a poor conversion rate, which suggests the content attracts casual viewers but does not build loyalty. Revenue data is also missing entirely. Best suited for creators who want to test high volume content strategies or explore underserved demographics without the time commitment of manual production.
Related Strategies
More AI agent strategies you might find useful
A LinkedIn Content Tool That Hit $62K MRR in Three Months
An AI LinkedIn content tool went from zero to $62K MRR in three months by solvin…
Running a Company Entirely With AI Agents: What Worked and What Broke
Seven AI agents ran a company on a single VPS for a week. They pivoted the busin…
12 n8n Workflows That Replaced $3,000 Per Month in Manual Work
Twelve n8n automations eliminated $3,000 per month in manual tasks covering lead…
Want more strategies like this?
Get weekly AI agent case studies in your inbox.