Vibecoding a data visualization dashboard for a Philippine health survey turned out to be “AI does the boring parts while you have to make a lot of editorial calls.” Jaemark Tordecilla covers the wins (data cleaning, pivoting fast, chart generation) and the walls he hit. (Generative AI in the Newsroom)
- A Philippine journalist spent a week vibe-coding a health survey dashboard using Claude Code, finding that AI handled data extraction and chart generation well but repeatedly failed on local administrative geography, placing cities in the ocean or rendering them like mismatched puzzle pieces.
- The "remaining 10 percent" problem was real: the last details, mostly ones requiring local civic knowledge, consumed disproportionate time and back-and-forth with the agent.
- The honest takeaway is that coding agents lower the floor for data journalism projects but don't replace the editorial judgment needed to decide what to visualize, how to structure it, and whether the output is actually correct.
