I started with a practical question: if someone has no customers, product, audience, or stable traffic, cannot code, does not want to appear on camera, has two hours a day, and can spend about RMB 1,000, which AI-enabled business direction should they test first?
Project 1: a map of possible directions
The first repository was XiaomingX's AI Money Maker Handbook. It collects four side-hustle principles, one starting strategy, ten technical monetization ideas, six technical business models, and twenty-five AI content directions.
I used it as a map. A beginner does not need to start with model training, a SaaS product, or a large investment. The more realistic path is to find a small demand, test it through public content, and keep reusable assets such as templates, scripts, images, and workflows.
The problem is that a map still contains too many roads. I needed a second project to compare those directions against one specific profile.
Project 2: a fit check based on real constraints
The second repository was muzhi-888's AI Side Hustle Checker.
I installed the Skill and entered the complete test profile:
- basic computer, search, AI chat, image, and editing skills;
- no programming experience;
- two hours available per day;
- RMB 1,000 in starting capital;
- no customer base, product, audience, private community, or stable traffic;
- no on-camera appearance;
- a preference for selling through public platforms instead of finding service clients first.
The first recording only showed the installation and did not capture all three outputs. I rejected it as incomplete evidence. A second uninterrupted 112-second recording captured installation, input, scoring, and the complete result.
The result
The project scored each route on cash-flow speed, repeatability, and payback period. Each category had a maximum of ten points, for a total of thirty.
- Niche digital products — 27/30
- Small, lightweight physical products — 24/30
- Content-assisted product sales — 22/30
The score measures fit with this test profile. It is not a success probability or revenue forecast.
The strongest route was a niche digital product: templates, checklists, spreadsheets, asset packs, printable files, or a focused information package for one concrete situation. These products are inexpensive to test, easy to revise, and can be sold repeatedly without inventory.
What this experiment actually proved
The test did not select a final product, and I will not pretend that it did. It narrowed a long list of industries into three product routes that match the stated constraints.
The next experiment will move from industry selection to product selection. I will use open-source projects to compare public demand, competing offers, keywords, production cost, and after-sales effort before choosing one specific digital or lightweight physical product.
Repositories:
- https://github.com/XiaomingX/ai-money-maker-handbook
- https://github.com/muzhi-888/ai-side-hustle-checker
Full experiment report and uninterrupted screen recording:
https://jianailab.com/reports/ai-side-hustle-checker-industry-selection
Top comments (0)