That Anxious Xiaomin

Last week, my friend Xiaomin (she runs Xiaohongshu account management for clients, pulling 20-30k RMB/month) met me for dinner, visibly stressed. AI helps her crank out 8 posts a day — more work keeps flooding in. But "should I raise my prices?" "should I take that big client?" "should I kill this line of work?" — those calls she still makes by gut. Result: a year of grinding, no more money. I got stuck in the same trap 2 years ago — pushing AI to write more, post more, take on more. Money didn't grow, anxiety did.

Foundry vs. Navigator — Where Do You Stand?

Silicon Valley's been talking about this split clearly: two types of players. Foundry: uses AI to crush the cost of "doing" — mass production, assembly lines, speed. In biotech, the examples are Insitro and Xaira — cranking out experimental data with high-throughput gear. Mapped to our small businesses: "AI spits out 50 images in one click," "batch-write 100 ad copies," "auto-reply to every DM." Navigator: uses AI to help you judge direction — take it or not, how much to charge, which lane to be in. In biotech, AI gets embedded into workflows as a decision consultant. Mapped to small business: "feed client conversations to AI, let it tell you if this client is worth taking"; "feed competitor info to AI, let it help you pick a niche." Most small bosses right now are Foundry players — cranking out with AI but nobody watching the road for you. Half a year later you'll find volume's up, direction's off, money didn't stay.

Trying Navigator Today — What's the Cost?

Money: 0-30 RMB/month. Existing ChatGPT, Claude, or domestic models all work. Time: 1-2 weeks to see results, need at least 5-10 real decisions fed to AI. Tech barrier: copy-paste into chat, no code. First step: tonight, grab a decision you've been waffling on (take this client or not / set a price / kill a project), paste all the context to AI, ask it for 3 options + pros/cons. You still make the final call. If you're short on judgment calls right now, stay in Foundry mode, run the volume, feed yourself first — this approach isn't for everyone right now.

Advice by Stage

- Just starting (no steady gigs yet): Foundry is fine for now. AI helps you mass-produce content, survive first, talk later. Navigator is your future stage, not now.
- Got 1-2 steady clients: start shifting to Navigator. Build the habit of saving every quote, retro, client message to AI. Weekly ask it "are these clients worth taking?" Half a year later you'll notice more money, not more work.
- Scaling up (5+ people): split Foundry and Navigator into two roles. One team owns delivery at speed (take fast). One team owns "take what / kill what / sell to whom" (pick right). The two teams balance each other — that's how you avoid wrecking.