Barbell Strategy for Entrepreneurs | Generated by AI

Home 2026.07

Question: My friend says I’m well-suited for entrepreneurship — that I’m someone who prefers high risk, but in practice chose a low-risk strategy. Thoughts on this exchange?

Answer:

Your friend nailed something real, and it’s not a contradiction — it’s the correct architecture. What you’re running is essentially a barbell strategy: high-risk ambition (Tinker-level depth, training models, betting on the agentic world) funded by low-risk operations (consulting revenue first, mortgage covered, 6-12 months runway). The risk lives in the upside, not in the survival layer.

This is also exactly how your idols did it. Liang Wenfeng funded DeepSeek from High-Flyer’s quant profits — cash-flowing business first, frontier research second. Karpathy built his reputation through teaching and open work long before Eureka Labs. The romantic “bet everything, burn the boats” founder is mostly survivorship bias; the ones who last tend to look boring on the spreadsheet and radical in the roadmap.

Your 种田/个体户 analogy is sharper than it sounds too. A one-person AI services company is the modern individual proprietor — except your “farm” scales in a way a restaurant never could. A restaurant owner’s output is capped by hours and square meters. Your leverage is code, models, and published knowledge that compound while you sleep. Same social position as the 个体户, completely different asymmetry. That’s the actual reason this era favors solo builders: the capital-intensive advantage big companies had (distribution, infrastructure, teams) is exactly what LLMs and cloud GPUs are eroding.

One thing worth holding onto from your friend’s framing: the danger for someone with your profile isn’t taking too much risk — it’s letting the low-risk service layer quietly become the whole business. The 20% product energy rule in your playbook is what keeps the barbell a barbell instead of just a safer job you gave yourself. Worth auditing every quarter: is the safe side still funding the risky side, or has it replaced it?


Back Donate