Things I've Done with AI
sjer.red
March 9, 2026
6 min read
🔥🔥🔥🔥🔥
48/100
Summary
The author began programming in middle school, initially creating HTML for a Neopets homepage. This evolved into developing static sites for Minecraft servers and Java plugins, leading to a career in engineering after an internship at AWS revealed the high earning potential in the field.
Key Takeaways
- The author has approximately seven years of professional programming experience and another seven years of self-taught programming as a hobby.
- The author initially hesitated to use AI tools like GitHub Copilot and Claude Code due to concerns about code quality and maintainability.
- The author shifted their perspective on programming, prioritizing problem-solving over writing aesthetically pleasing code, and has not written traditional code since October 2025, focusing instead on prompts and reviewing AI output.
- In the past nine months, the author has completed multiple projects using AI tools, including migrating from Jenkins to GitHub Actions and Dagger.
Community Sentiment
Positives
- Using AI tools like Claude and Gemini 2.5 Flash has significantly improved my understanding of complex tax situations, showcasing AI's potential for practical applications in personal finance.
- The ability to automate tedious tasks with AI, such as organizing and extracting information from tax documents, demonstrates how these tools can enhance productivity and efficiency.
Concerns
- There is a growing concern that AI tools may lead to obsolescence for users, raising questions about the long-term impact of relying on these technologies for productivity.
- Critics argue that many AI projects focus on superficial improvements rather than meaningful productivity gains, suggesting a potential disconnect between AI capabilities and user needs.
