[Advice Request] Seeking Guidance: How a Junior Software Engineer Survives and Rebuilds in the AI Era

Dear Senior Colleague,

I hope this email finds you well. First of all, thank you for your
incredible contributions to the field of computer science; your
engineering practices and insights have deeply inspired me.

My name is Huijian Qin. I graduated in 2025 with a Bachelor's degree
in Artificial Intelligence from Xiangtan University, a "Double
First-Class" university in China. After graduation, I worked as a Java
Development Engineer at a research institute, and later as a Software
Engineer at a robotics company. However, I recently made a tough
decision: I resigned.

The reason behind this stems from a profound sense of conflict I
experienced in the workplace. Facing the AI wave, I initially insisted
on hand-writing code to build my foundation, but my output couldn't
meet the company's efficiency demands. When I pivoted to heavily
relying on AI, I faced a new nightmare: dealing with unmaintainable
AI-generated "spaghetti code." More frighteningly, I felt my
foundational coding intuition and system design skills rapidly
degrading. With veteran developers criticizing our generation's
reliance on AI, I fell into an exhausting cycle of using AI during the
day to survive deadlines, and staying up late to hand-write code just
to maintain my basics.

I realized that despite my work experience, I haven't yet built the
systematic capability to independently solve complex engineering
problems. Therefore, I decided to hit pause. I will be spending the
next 5 months in an intensive, full-time program at "Itcast" (a
well-known IT training bootcamp in China). My ultimate objective is
clear: to systematically rebuild my problem-solving abilities so I can
truly survive and thrive as a Software Engineer in the AI era. (Once I
stabilize, I also plan to pursue an online Master's degree in CS from
an overseas university).

To complement this 5-month systematic rebuild, I am currently adopting
a "dual-track" strategy for my daily practice and personal projects.
As a junior developer, I would love to hear your thoughts on this
approach:

Acting as an AI Reviewer (Tech Lead Mode): For projects, I let AI
generate the actual code. I force myself to act like a "Tech Lead,"
focusing strictly on two things: giving precise instructions (prompts)
and rigorously reviewing the AI's code and underlying logic for
correctness and robustness.

Hardcoding Fundamentals (Feynman Technique): Since the first approach
only trains my logical thinking and not my coding muscle, I compensate
on the other end. I practice algorithms daily (LeetCode) and read
classic CS books. Furthermore, I strictly apply the Feynman Technique
to break down, explain, and deeply internalize these core concepts.

As a junior developer trying to navigate this massive industry shift,
I would love to ask for your perspective:

Do you think my strategy of "using AI for implementation while acting
as a reviewer, combined with grinding LeetCode and the Feynman
Technique for fundamentals" is the right direction?

In an era where AI can handle most basic coding tasks, what is the
true "moat" a junior software engineer must build to survive
long-term?

I know you have an incredibly busy schedule, so even a one-sentence
reply or a quick piece of advice would be immensely valuable to me.
Thank you so much for your time.

Best regards,

Huijian Qin