Over the past few weeks, I've been collaborating with AI tools, including ChatGPT and Cursor, on one of several ongoing projects designed to deepen my understanding of AI in programming. This particular project focused on building a personal website, drafting case studies, and experimenting with how AI can assist in both technical execution and content creation.
The primary goal was twofold: to explore the strengths and weaknesses of AI in coding tasks, and to refine my own ability to craft functional, precise prompts that guide AI effectively. This process highlighted a few key insights:
AI as a collaborator, not a replacement
AI tools excel at generating boilerplate code, suggesting structure, and providing rapid content drafts. However, human oversight is essential for context, correctness, and aligning outputs with specific goals.
Prompt engineering is a skill
The quality of output is highly dependent on how clearly the task is described. Iterating on prompts and refining instructions became as important as reviewing the AI's outputs.
Learning by doing
Using AI to scaffold the website and draft case studies provided hands-on experience with both the potential and current limitations of AI-assisted development. The process reinforced the value of iterative collaboration — combining AI's speed with human judgment.
High-level planning remains critical
Even with AI support, understanding the architecture of the site, the flow of content, and the strategic presentation of case studies required deliberate design decisions that AI alone cannot make.
This project is part of a broader series of initiatives aimed at improving my understanding of AI in development. Each project provides new insights, challenges assumptions, and strengthens the practical skills needed to leverage AI effectively.