Skip to content
Lucky Snail Logo Lucky Snail
中文

2024 Musings

/ 6 min read /
#年度总结
Table of Contents 目录

Hi everyone, I’m luckySnail. By the time I’m writing this, it’s already the fifth day of 2025. Looking back at the past year from this point, though there were shortcomings and regrets, overall I’m still moving forward. I’m used to writing technical articles where the words flow naturally, but suddenly writing something without a theme, I found myself staring at a blank page not knowing what to say. So I’ll just write whatever comes to mind!

How AI Changed Me

Over the past year, the place where I typed the most wasn’t a code editor or the WeChat input box — it was the AI conversation input box. I’ve truly become a developer heavily reliant on AI.

In the past year, I’ve written at least 10 articles about AI. In my opinion, AI coding has become incredibly powerful, all thanks to Claude 3.5 Sonnet and Cursor. If you’re not familiar, check out my article: https://www.luckysnail.cn/posts/post-27 . If you want to dive deeper into this field, I recommend following “Baoyu” (宝玉), his personal blog: https://baoyu.io/ .

Since models are now very capable, I think we can get decent results without highly professional prompts. The key is to keep experimenting and accumulating experience. Here are some tips I’ve compiled: https://www.luckysnail.cn/posts/post-31 . Baoyu also shared his own experience: https://baoyu.io/blog/2025-prompt-engineering-tips .

Over the past year, the AI tools I’ve used include:

Among the tools above, I’m currently using Devv and GitHub Copilot. Devv aggregates most of the strongest LLMs and I think it excels in RAG and interaction. GitHub Copilot is used alongside VS Code — it currently offers free quotas and supports a wide range of capabilities, which is sufficient for me.

AI is definitely a boon for programmers. It enables an average person like me to quickly turn ideas into reality, rapidly learn new knowledge, and see the hope of becoming a super individual.

In the new year, I plan to spend less attention on trying out new AI tools — though I’ll still pay attention if something really stands out.

Cognition

2024 was undoubtedly the year I gained a clearer understanding of myself and acutely realized my cognitive shortcomings. This year, I superficially learned that:

  • Completing development is just the beginning; the real heavy lifting is in maintenance after launch, regression testing, and so on.
  • Code review is important: writing the code is only the start; you still need to think about whether it’s the best implementation, whether there are edge cases not covered by the business, etc.
  • Technology is merely a means to solve problems.
  • Good design can reduce later communication with customers; bad design leads to nothing but feedback.
  • Feature implementation matters, but usability is equally important.
  • Good communication can reduce a lot of work.
  • Pay attention to data, to customers, to business value.
  • The world is a huge makeshift operation; life is a wilderness.
  • Passion is the primary productive force — passion.
  • And many, many more…

For many of the above, I haven’t truly integrated knowledge and action. I’ve only briefly known them. I’m writing them down now. In the new year, I want to live and work with a more humble and peaceful mindset.

Attention

This year, I experienced a very famous syndrome: FOMO (Fear of Missing Out). I found that I’m also seriously affected. 2024 was a year hijacked by useless information, which only added unnecessary worry and anxiety.

This year, though I spent a lot of time on entertainment, I was more disciplined compared to my past self. In the new year, I’ll be satisfied as long as I’m a little better than 2024.

This year, besides consuming useless information and entertainment, I did learn new things as well.

I hope in the new year to pay more attention to myself and be more relaxed.

I hope that in the new year, besides age +1, other aspects can also increase, increase, increase.

Learning

This year, even though I’m usually thrifty, I spent a few thousand RMB on knowledge — mainly on:

  • Programming navigation and frontend vertical knowledge planets: personally, I see these more as paying for community.
  • AI coding: I reluctantly bought the Devv annual subscription — I think it’s one of the most cost-effective paid items of 2024.
  • Structured learning courses: I bought some second-hand courses and also purchased some Nuggets Booklets. I understand these as someone else spending time to organize a learning path for me, and I follow it step by step.

This year, I relied less on video learning, finding it relatively inefficient. I tried AI + official docs (or curated learning paths) to learn, and the results have been better so far. Video tutorials have significant downsides:

  • Uncertainty: during learning, the video author might go off on tangents, leading to extra knowledge that you might not use, effectively meaning you didn’t learn it.
  • Mental burden: typically, learning tutorials are quite long, causing psychological pressure.

This year, I practiced the Feynman Technique more often. While learning, I would deliberately write technical articles about important but easily forgotten knowledge to deepen my memory. Even if I forget later, I can look it up. In practice, after learning Docker, I actually had a need to add a Dockerfile to a project — my previous notes were a huge help.

This year, I also paid more attention to learning methodologies. I recommend two articles:

  1. https://mp.weixin.qq.com/s/6vAzb0Vl9-KzFG10ceBrSw
  2. http://blog.yuanjin.tech/article/103

They’re worth re-reading. I haven’t absorbed and turned many of the insights into actual actions.

That’s my rambling about learning this past year.

Life

In terms of life, 2024 brought me many new experiences.

  • I went to see Liang Bo’s concert (a multi-artist show).
  • I went to Anji for water rafting.
  • I experienced a three-month fitness journey.
  • I did my first external sharing — although very nervous, I took the first step.
  • I was responsible for a project for the first time. Looking back, I did a terrible job, but it gave me a clearer understanding of myself.
  • This year, music remained my favorite hobby, occupying the most time.
  • I re-watched the TV series Blossoms (繁花) again — I was once again blown away by Wong Kar-wai. If you also like it, they recently released a Long Time No See anniversary short film — worth watching.
  • And a few others…

For 2025, I hope life can be stable and calm, and with limited energy, I want to experience more new things.

The Road Ahead

In life, I hope to become someone who can be more flexible and adaptable.

At work, I hope to be able to stand on my own and gain more trust.

In one sentence: just do it.

2025 Flags

More holistic goals

Work & Professional Skill Improvement

  • Be more meticulous: requirement completion rate 99%+, production bugs no more than 5, no low-level mistakes.
  • Complete React source code learning and implement a snailRun version of mini-react.
  • Perform better at work and communicate more efficiently.
  • Organize daily work and create my own workflow.
  • From frontend => full-stack.
  • And a few others…

Life

  • Experience new things.
  • Read books.
  • Continue sharing content about work, life, tech, and AI.
  • Analyze the time spent on daily routines and turn repetitive tasks into workflows.
  • And a few others…