跪拜 Guibai
← All articles
Frontend · Backend · Product Manager

A Year Offline: Betting on AI, Becoming a Dad, and Going Back to School at 34

By 我是Allen ·
Read original on juejin.cn ↗ Google Translate ↗ Alt translation

The post captures the real cost of an AI pivot inside a SaaS company: constant tool churn that erodes technical depth while sharpening customer-facing skills. It also surfaces the credential pressure that still shapes career decisions in China's tech market, where a degree can gatekeep roles that experience alone cannot unlock.

Summary

A year away from writing reveals a developer stretched across three life-changing fronts. His company's SaaS business went all-in on AI, pulling him into a relentless cycle of tool-switching — from Copilot to DeepSeek — where model updates constantly invalidated hard-won experience. Technical depth stalled at the application layer, but soft skills hardened: translating customer pain points without jargon and building systematic knowledge of vertical industries. At home, a pregnancy reshaped every daily decision, trading personal hobbies and a social life for a healthy baby and a brief, low-revenue experiment in the maternal-infant content track on Douyin and Xiaohongshu. Simultaneously, a blunt warning from an HR professional — "your educational background is a hard flaw" — pushed him into a part-time graduate program. The daily rhythm became dinner, one hour with the baby, then books and notes until late, sustained by a simple rule: if you're not smart, just put in more time.

Takeaways
An entire SaaS team pivoted to AI, forcing a year-long cycle of adopting and discarding tools like GitHub Copilot, Kimi, DeepSeek, and Claude Code as models iterated.
Technical depth suffered because most work stayed at the AI application layer — understanding customer needs and wiring AI into business scenarios — with little access to framework or core logic development.
Soft-skill gains included translating customer pain points without technical jargon and building systematic knowledge of several vertical industries' business processes.
Having a child triggered a concrete reallocation of time, money, and energy: hobbies stopped, social circles shrank, and daily life became a negotiation over meals, prenatal visits, and childcare.
The family briefly entered the maternal-infant content track on Douyin and Xiaohongshu, using their baby as a child model, though the income was negligible.
An HR professional's comment that his educational background was a "hard flaw" drove the decision to pursue a part-time graduate degree at age 34.
The study routine was rigid: after work, dinner, and one hour with the baby, he sat at a desk reading, doing exercises, and taking notes every night for a year.
Receiving the admission letter brought mixed feelings — joy alongside the weight of compressed family time and significant tuition costs.
Conclusions

Tool churn in AI is not just a productivity tax; it actively prevents depth by resetting expertise before it compounds, leaving engineers stuck at the integration surface.

The HR remark about a "hard flaw" in educational background reveals a hiring reality where credentials remain a hard filter, even for experienced engineers, pushing mid-career developers back into formal education.

Treating a baby as a child model and launching maternal-infant social accounts turns a life event into a low-stakes entrepreneurial experiment — a pattern of side-hustle thinking that surfaces even during personal upheaval.

From the discussion

The discussion splits over career strategy: one view questions why someone capable of getting into a top university wouldn't aim for a stable civil service job, while a counterpoint argues that the civil service exam is now harder than graduate school entry. A reply resolves the tension by suggesting a graduate degree makes the civil service exam easier, framing the author's path as a smart long-term play. Other comments are simple praise or requests for more detail on the author's background.

Getting into a top university like UESTC in one year at over 30 demonstrates high ability, raising the question of why that effort wasn't directed at the civil service exam.
The civil service exam is currently more difficult than the graduate school entrance exam, challenging the assumption that the author took an easier route.
A graduate degree can make the civil service exam easier later, positioning the author's choice as a strategic, two-step career move.
Some readers simply express admiration for the author's achievements.
Curiosity exists about the author's prior education and specific field of graduate study.
Featured comments
yuanchen 1 likes

At over 30, you prepped for just one year and got into UESTC. With that kind of ability, why not take the civil service exam? [thinking emoji]

ryanliu

These days, the civil service exam is way harder than the grad school entrance exam.

_soga  → ryanliu

Once you've got a grad degree, the civil service exam becomes easier.

See top comments, translated →
Source: juejin.cn ↗ Google Translate ↗ Backup ↗